My onetime professor Lawrence Solum spots several ambiguities in this formulation, and points out that non-originalist theories - common-law constitutionalism, pluralism - don't necessarily produce Brown, but only are amenable to Brown. I will assume for purposes of this post that Dorf means that, if the Court accurately applied originalism in 1954 to the facts in Brown, it must be the case that there's a significant possibility that the Court would have reached the result it actually reached - as I agree that Dorf probably can't be demanding a theory that would invariably generate the result in Brown, and because it appears to me that his argument is that originalism, done right, never produces the result in Brown as an original matter.
Assuming that's what Dorf means (if I'm wrong, it's nevertheless something worth responding to, as it's a reasonable-enough claim that others have made), I see two problems. First, just why is it that originalism must be capable of generating Brown as an original matter, rather than as a matter of stare decisis were the Court to adopt originalism in 2015? Second, even if we do grade originalism on how it would perform in 1954, why is it even that important that originalism be capable of generating sacrosanct precedents at all?
On the first question, I take it that when originalists call for the courts to adopt originalism, they propose a reform in interpretive methodologies going forward (unless they believe the courts have already adopted originalism) - not an alteration of past interpretive methodology by means of time machine. So since Brown is the law, and since most originalists believe in stare decisis (in part because stare decisis existed in 1787 and was part of the package of original understandings of how Article III courts would behave), originalists who call for originalism going forward call for adherence to Brown, a decision that doesn't demand to be overruled under any form of stare decisis. Why does it matter, then, that, if originalism had existed in 1954 and originalists had controlled the Court in 1954, we couldn't (maybe) have gotten Brown? Dorf says it's because past performance is a predictor of the quality of future results:
Consider an analogy. Suppose I'm trying to decide which of two builders to hire to build a house: Stubbs or O'Reilly. Each of them shows me several houses he has previously built. The Stubbs houses are high quality but the O'Reilly houses are falling down. O'Reilly nonetheless wants me to hire him. I'm skeptical because I see what lousy work he has done in the past. He tells me not to worry because he himself lives in a high-quality house that Stubbs built. O'Reilly is proud of the fact that he hasn't destroyed the excellent house Stubbs built. But I'd be a fool to hire O'Reilly because the capacity not to destroy an excellent house is not a predictor of the ability to build an excellent house.This analogy doesn't really work. Originalism isn't just capable of not destroying excellent houses; it can build some excellent houses of its own. No non-originalist would claim that the entire original meaning of the Constitution in every particular is unworthy of adherence. Even liberal anti-originalist pundits like Ian Millhiser are discovering that they like what original meaning has to say about birthright citizenship; any number of thoughtful liberal non-originalists like what they believe original meaning has to say about affirmative action (see pages 81-85). And people of all ideologies find a lot to like about the Bill of Rights or divided government, even as originally understood.
Originalism supplemented by stare decisis is like O'Reilly.
So what we need is to ask whether a past inability to generate one excellent decision is a predictor of an inability to generate certain other (or enough) excellent decisions going forward, though we know originalism is sometimes capable of generating excellent decisions. What we might find if we asked that question is that, while originalism was not equal to the challenges posed by Jim Crow, or possibly the Great Depression, originalism can produce excellent and even sacrosanct decisions that dispose of the outstanding constitutional questions that arise today. This, of course, is highly contingent on the mix of questions that arise, what originalism has to say about them, and what one's idea of a sacrosanct decision is in the first place. I would expect, for example, that if someone like Trump were to get elected President, a non-originalist Court would give his arguments against birthright citizenship a much more sympathetic hearing than an originalist Court would, given the strength of the originalist arguments for birthright citizenship and the tendency of this non-originalist Court to take most any constitutional claim seriously so long as it's backed by one major political party and a large share of the public. Whether a decision upholding birthright citizenship would be sacrosanct would be in the eye of the commentator, but I imagine the bulk of the people who decide whether decisions are canonical or not would quickly canonize such a decision.
On the second question, suppose that for some reason we did care what originalism would have done in 1954, not just what it would do in 2015. Is it really a devastating objection to originalism that originalism would have failed to produce Brown? I don't see why. The argument that it is seems to go something like this: (1) Segregation was a horrible evil. (2) Brown (in theory) ended segregation. (3) Therefore, Brown is a great and sacrosanct decision. (4) We would want any interpretive theory to be able to produce Brown.
My trouble with this argument is steps three and four. It's obviously the case that Brown was an immensely welfare-enhancing and virtuous decision, because it helped end a horrible evil. But is any immensely welfare-enhancing and virtuous decision sacrosanct, and must a good interpretive methodology be capable of generating certain decisions that generate immensely welfare-enhancing social change? What's wrong with acknowledging that the Constitution just doesn't say certain things we'd like it to say?
For example, slavery was an even greater evil than segregation. The Constitution, however, permitted slavery until it was amended to forbid it. In a counterfactual world where the Thirteenth Amendment was never ratified, we could imagine a non-originalist decision that read the Constitution to ban slavery. Would we say that originalism was defective because it couldn't produce such a decision? I think not. It doesn't follow from the immense virtue of some social/legal change that the Constitution must be capable, in its current unamended state, of generating that change by means of judicial review. It only follows that it is morally imperative that the change occur by some means, which certainly need not be judicial. Further, I would think that this moral imperative runs most directly to political branches that can change the law, rather than to courts interpreting the old law they're stuck with, which may well be flawed from a moral perspective. That is, there was obviously a moral imperative in 1954 to pass laws ending segregation, or to amend the Constitution to do so, which comes to the same thing. But was there really a moral imperative, failing those developments, to read a constitutional provision enacted in 1866 to ban segregation? Perhaps if segregation was so popular that Brown was the only way to end it, but I doubt that's the case; rather, to crudely simplify things, Brown only was enforced once segregation became unpopular enough for political actors to enforce Brown.
Take another example. It may be quite unjust that poor children are penalized for the poverty of their parents and neighbors and go to public schools that receive less funding than those attended by children fortunate enough to be wealthy. It would be nice if the Constitution had something to say about this. But is it a black mark against originalism, or the non-originalist methodologies actually used by the Court that rejected poor children's constitutional claims in 1973, that they respectively cannot and failed to generate a decision that rectifies that injustice? No more so, I suppose, than it's a black mark against our state constitutions that they produce state legislatures which invariably vote to maintain the same unjust system of property taxation and school funding - a fact which leads very few people to say that our state constitutions embody a morally infirm system of government which must be replaced by some other system of government.
For whatever it's worth, I am not an originalist and thus happily embrace Brown as an original matter; I have no idea what I would say about Brown were I an originalist. And I do think it's a selling point of certain non-originalist methodologies that they're more amenable to Brown and other canonical decisions than originalism is. But unless one believes that the Constitution just is an unerring fount of moral goodness, which seems doubtful, or that even if it isn't, an interpretive methodology must make it so, which also seems doubtful, I don't see why an interpretive methodology must be capable of generating every profoundly morally good decision in the constitutional canon.