When Legal Realism Gets Internalized

Back in the old days, common law judges had what we would now consider a naively romantic view of their jobs. They did not see themselves as making the law, but rather as using the principles of Reason to "discover" the law. The law, they thought, had external validity. It just existed, out there, somewhere, sort of as natural law was thought to. Some of them might even have said that the common law they applied was natural law, though others might have denied it. Then, starting a little more than a century ago in this country, came the legal realists, who basically demolished this idea. They pointed out, irrefutably so, that an awful lot of judicial decision-making was influenced by the personal characteristics of individual jurists, the politics, the identities, the ideologies, etc. That was the empirical claim. It got kind of wrapped up in a philosophical claim about what law is. To a legal realist the idea of the law's existing independently of the legal system was absurd. Put perhaps a bit too concisely, the law was seen as nothing more than a prediction about the behavior of judges.

This story, of the rise of legal realism and the corresponding fall, and then demise, of the classical view of law, kept occurring to me earlier tonight as I read the line of Supreme Court cases about federal common law. Here the story is as follows: in Swift v. Tyson (1842), Justice Joseph Story declared that, in a case which only gets into federal court because the parties to it are from different states, i.e. where the legal issues are not ones of federal but rather of state law, the federal courts were not bound to follow state common law. Part of the motivation may have been the desire to create a uniform national commercial law through the federal common law. In any event, this was the rule for very nearly a century, until the Court overturned Swift in Erie Railroad Co. v. Tompkins (1938). And I just couldn't shake the feeling that the difference between the two cases is the difference between classicism and realism. Moreover, though I'm not at all sure which of the two cases I think is better, it seems to me that Erie and its progeny showcase the potential dangers of letting the legal system internalize even the most valid insights of the legal realists.


If you believe the common law exists independent of judicial decisions, and that these decisions do not make law but merely discover it, Swift v. Tyson made an awful lot of sense. The question of what constituted negligence had, in this view, some answer, which was the right answer regardless of what any court said the answer was. Why, then, would a federal court just accept the answer of some state court? The state court might have gotten it wrong. Indeed, if there's anything to debate about which approach to take, the federal court necessarily does think the state court got it wrong. In these diversity cases the statutory laws of the relevant state governed, that was beyond doubt, and state court decisions interpreting them governed as well, but when the state deigned to pronounce judgments on no basis other than the logic underlying all the common law, well, the federal courts were under no obligation to agree with their logic. There was only one common law in principle, and there should be a distinct "federal common law" precisely because federal courts should apply the common law in the way they thought was right.

This gets awfully problematic once we let the legal realists into the room and they tell us that this independent common law doesn't really exist. Because then the claim of federal courts that they must apply that independent, unwritten law as best they saw fit just kind of... disappears. Moreover, to a legal realist there's not much difference between judge-made law and statutory law. To an individual in society, and particularly to Oliver Wendell Holmes' "bad man," the key question is the same in both cases: what will actually happen if I get hauled into court? It is no matter of the federal government's which institutions a state uses to expound its laws; indeed, not every state has a common law, and it is in principle no business of the feds to care about that. If state law is to govern in diversity-of-citizenship cases, and if state common law forms a part of the laws of the state just as much as statutory law, then federal courts must follow state common law.

Well, we know that the legal realists are right, at least in their battles with what had come before. So isn't Erie right? Well, in a sense, and since the two cases are on one level a matter of statutory interpretation it may well be true that Erie is the better approach. (Although the Court in Erie does seem to suggest that federal common law in the Swift sense would be unconstitutional, somehow, which is strange...) But it is curious, I think, that this legal realist philosophy is informing this decision of the highest court in the land. For everyone who isn't a judge (or perhaps a legislator), the idea that law is nothing but a prediction of the behavior of courts is a fine one. That idea actually forms a significant part in my own thinking about the concept of "judicial supremacy." But can a court believe that, and still function? The court's job is to apply the law to the case before it, but if it thinks that "the law" is nothing more than a prediction about what it will do, well... what next? You could perhaps derive an extreme form of stare decisis from that mindset, and say that no court should ever render a ruling which the best legal experts would not have expected, but no one really thinks that. So what is a legal realist court to do? Whatever it does will, retroactively, have been "the law," so can't it do whatever it wants?

Or consider a development that came right after Erie. It was recognized, particularly given the new Federal Rules of Civil Procedure, that federal courts were not obliged to behave exactly like state courts. They did not have to pretend to literally be extra state courts, when hearing diversity-of-citizenship cases, and the procedures by which they operated could remain distinctively federal in those cases. But where lay the line between substance and procedure? For example, which were state statutes of limitation? A statute of limitation says that a certain right may not be pursued in state court after a certain length of time. Must federal courts respect these limits? From a legal realist perspective, of course. If the ability to seek relief to enforce a certain right expires after a certain date, the right does as well. After all, what is a "right" which cannot be enforced in court? Not law, that's for sure, since its existence tells us nothing about what the courts will do. So if a plaintiff could not have sought relief in state court due to a statute of limitations, it may not seek that relief in a federal court either.

But this is troubling. Consider one of the parts of Marbury v. Madison that I love, and not one of the famous parts. Chief Justice Marshall's opinion has several stages. First he asks whether Marbury was entitled to his commission as Justice of the Peace for the District of Columbia, and he answers, dubiously, that he was. Then comes the part I care about for these purposes, where he asks whether there is any remedy Marbury can seek to get that commission. There is, Marshall concludes, because where there is a right there must be a remedy. Essentially he is making the opposite inference from the legal-realist approach to statutes of limitation. They argue that the lack of a remedy must imply the lack of a right; he says that a right must imply a remedy. Technically they agree! Those inferences are the contrapositive of each other, and both say that right and remedy must exist together or not at all. But because we don't live in formal logic land, which direction of the causal arrow we emphasize has genuine consequence.

And this, I think, gets to the problem with legal realism, as internalized by the legal system which it critiques. A legal system which takes a realist view of itself must, I contend, sacrifice any internally aspirational elements. It must forget about the celestial city that my grandfather described in his essay "The Two Cities of the Law." It must abandon the idea that there is any such thing as "the law as it ought to be," distinct from "the law how it is." Merely accepting that courts can't go crazy with natural law, that "the law how it ought to be" doesn't become "the law how it is" by the simple virtue of its being right, doesn't commit you to this sacrifice. Internal legal realism does. Under a suitably extreme legal realist doctrine, any claim of a right not currently recognized by the courts must fail. After all, how can there be a right where courts are not providing a remedy?

Now, legal realists would not argue that the law should not be improved. They would just say, however, that the improvement could not be expected to come from within the law. And they might be right, as a pragmatic matter. But that doesn't mean we shouldn't ask courts to try their best to improve the law. Maybe they won't do it, or maybe they'll do it wrong, and maybe in the end we can only count on the political process to make the world better, although the several decades right after Erie was decided seem to vitiate that notion. But it would be nice if they would at least keep in mind that "the law as it should be" is out there, and though it can't just become positive law, we should be reluctant to pass up opportunities to make the positive law a better approximation of it. It would be nice if the law would keep in mind that, though it cannot and should not live through logic alone, that does not mean it cannot try to live by principle. It would be nice if the law, acting purely within its own sphere, free from external influence, would keep the commitment to justice in the forefront of its mind. And it would be nice if the law followed John Marshall and insisted on finding ways to remedy the plight of those who have been wronged, rather than doing the easy realist thing and using the lack of an existing remedy as an excuse to ignore the wrong. Legal realism, in other words, is a great thing from an external perspective, and courts should keep in mind that their old mindset has been pretty thoroughly debunked. But it would be very nice if they wouldn't just adopt the realist philosophy as their own, but rather preserved the chief virtue of the old ways: its aspiration to be right.