Multifactor Tests Are Everywhere and Nowhere

A narrative explanation for multifactor tests?

The law seems to be filled with multifactor tests. There is debate about balancing, with some judges showing a “fear of balancing,” and some scholars urging that balancing is “inevitable.” But all that worry and guilt about balancing has not changed the fact that the legal form that so often requires balancing by judges—the multifactor test—seems to be going strong.

The multifactor test appears to date to the middle of the twentieth century. An article by Mitchell Johnston suggests the 1960s was the key date for the emergence of multifactor tests that have a series of steps. I found a similar phenomenon with the four-factor test for the preliminary injunction (p. 827 here). And now they are everywhere, in substantive law, procedural law, and remedial law. Like the great article title by Gary Lawson, we seem to be faced with “the rise and rise” of multifactor tests.

But there is something odd about this apparent success of multifactor tests—success, that is, in a Darwinian sense. What does the existence of the multiple factors actually do?

The plaintiff says that all the factors favor her. And the defendant says that all the factor favor her. By itself, that polarized pairing might not seem so odd, but then there’s this: the court decides the case, concluding that all the factors favor one party or the other. Sure, there are occasional exceptions. But they are rare. Usually all the factors line up. Yet the factors are supposed to have some independence, or else what’s the point of having a multifactor test?

In trademark law, two decades ago Barton Beebe found that “judges stampede specific factor outcomes to conform to or support the overall test outcome. The data suggest that judges determine the test outcome based on a limited number of core factors and then adjust the rest of the factor outcomes to accord with that result.” Thus, in Beebe’s telling, “This represents strong evidence of coherence-based reasoning in the courts.”

Two decades earlier still, Dick Fallon offered a “constructivist coherence theory” for interpreting the Constitution. He noted the phenomenon that for an interpreter of the Constitution in any particular instance, all of the familiar kinds of constitutional argument—text, framers’ intent, larger purpose, precedent, and justice or social policy—will tend to point in the same direction.¹ As Fallon put it, “within our legal culture, it is the rare judicial opinion, the anomalous brief, the unusual scholarly analysis that describes the relevant kinds of arguments as pointing in different directions.” To be sure, Fallon was not describing a multifactor test per se—rather, a set of multiple kinds of argument, which do not have a clear order of operations. Even the most dominant argument, the argument from constitutional text—the one argument on which every successful constitutional litigant must at least fight to a draw—does not have to be the first one considered by a court.

More recently, I have diagnosed a collapse of the preliminary injunction test, with the single factor of the prediction of the merits often deciding whether a preliminary injunction will or will not be granted (see here and here). Which is not good, since the point of the PI is not to decide the merits, but to preserve the court’s ability to later decide the merits. To be sure, however, that collapse has not been total, and there are decisions from the Supreme Court that emphasize the provisional and jurisdiction-saving quality of the preliminary injunction (see, e.g., Lackey v. Stinnie).

So we seem to be in a long phase of multifactor decadence, where multifactor tests and their fellow travelers seem to be everywhere and nowhere; they seem to be formally present but not really doing much operational work. (I reserve judgment on how much multifactor tests used to do work; I will say that for preliminary injunctions, the equities used to be more important than the merits prediction.)

This seems to be an existential failure: it’s supposed to be multifactor, but it turns out to be a kind of unifactor test. What’s happening?

One answer is that the factors are analytically interconnected, and that the reasoning process is iterative, with the factors gradually converging in a deliberative equilibrium. This is Fallon’s account, and he emphasizes the interconnectedness of the types of constitutional argument with respect to values.

Although there is force to Fallon’s argument, I don’t think it is a full explanation. For one thing, the relative strength of the different constitutional arguments has shifted since the 1980s, and arguments drawing on history have waxed as some others have waned. An article describing our practice of constitutional interpretation today would probably not, as Fallon’s article did, use cognates of value 255 times.

For another thing, a lot of multifactor tests don’t seem to rely on an iterated process of interpretive deliberation. When Fallon wrote his article on constructivist coherence interpretation, it was thirty years before the shadow/interim docket became so central to the Court’s work. Criticisms of that docket will fault, among other things, the lack of deliberation. Whenever someone endorses a particular action by the Court on the interim docket, the argument tends to be not that there’s been so much deliberation, but rather that an emergency situation called for the Court to act without the ordinary deliberation. Justified or not, any particular action by the Court on the interim docket does not look much like Fallon’s process of consideration and reconsideration, with reciprocal influence of the arguments and iterated reassessment of their force. At a minimum, Fallon’s analysis does not fit what is happening with stays of preliminary injunctions.

Amid this multifactor malaise, perhaps there is another way to think about what multifactor tests do. In our adversarial system, the plaintiff needs to tell a story and the defendant needs to tell a story. What the multifactor test does is establish what plot points the story must contain. Tell your story and make sure you hit all these points. You can’t leave any of them out. But it is the story that persuades the judge or doesn’t persuade the judge. That is, the persuasion happens at the level of the story as a whole, not the plot point. And then, persuaded by one story or the other, the judge then writes out that story, hitting all the same plot points required by the multifactor test.

I don’t mean to suggest this is how all multifactor tests work all the time. Nor do I mean to suggest an unvarying approval or unvarying disapproval of this kind of narratival function for the multifactor test. But I am suggesting, descriptively, that the tests can work this way.

Even when this narrative explanation fits, we still don’t want stories with just a single plot point. If the multifactor test is going to do any work, the plot points should not collapse together. But then, even when the plot points are independent, it shouldn’t be surprising when courts usually find that all the factors wind up pointing in the same direction.

1. With some refinement, Fallon’s kinds of constitutional argument, which approximate Philip Bobbitt’s modalities of constitutional argument, are an organizing theme in the Paulsen-McConnell-Bray-Baude casebook on constitutional law.