The Randomized Hypothetical
Imagine that we could construct the following system: Instead of a constitutional court or supreme court with constitutional judicial review, we design a randomized system of judicial review.
Here is how it might work (hypothetically, of course): Every new law enacted by the legislature is automatically submitted to the “judicial review computer.” Similarly, every constitutional challenge to a governmental policy or practice is filed with the same computer system (instead of the courts). Let us assume that a panel of lawyers feeds the computer with the set of possible legal outcomes of each challenge. Normally the set would be either pass or fail constitutional muster, but sometimes it could be a bit more complex, perhaps dividing the challenge to several options. As a simplifying assumption for now, we will postulate that the set of outcome options is both very limited in scope and fairly techni- cal.[4] Then, at the end of the year, the computer runs a program that yields a totally random selection of “cases” that it strikes down as “unconstitutional” and therefore legally invalid. How many of them? Well, we can easily determine some formula in advance, say, a certain number of cases based on factual parameters gleaned from the history of judicial review in the last century or so[5] - or any such mechanical, but essentially randomized, method. Let me call this the Randomized Judicial Review process, or RJR, in contrast with the actual Constitutional Judicial Review system we have, which I will henceforth label as CJR.Obviously, the RJR system would have to be a bit more sophisticated and complex for it to be plausible, even as a hypothetical. For one thing, we would need some initial screening procedures. For another, we would need some process, judicial or other, to determine some basic factual findings that would be needed to ground the constitutional challenges.
Both of these issues can be resolved, however, without insurmountable difficulties. We can imagine a system whereby lower courts would have to certify constitutional challenges, and determine their factual groundings, before they can be filed with the randomizing computer.So here is the question I would like to pursue in this essay: How would RJR differ, in significant moral-political ways, from CJR? I will try to show that between the hypothetical RJR and the actual CJR the differences are rather insignificant, morally speaking, and, in any case, provide no good reasons to prefer the actual to the hypothetical. At this point you might think that the issue is moot because there is absolutely nothing to support a randomized process of judicial review; it is just too crazy. Well, crazy it might be, but two considerations lend it some support: First, just like the current CJR, it puts a limit on majority rule. It curtails, to the same extent, at least quantitatively (ex hypothesis), the majority’s ability to enact laws or implement policies by a regular majority vote. Second, a consideration of fairness may count in favor of a purely randomized system. When you have a winner and a loser in a legal battle, and neither side is obviously right or wrong (more on this later), a randomized decision procedure gives each side an equal chance of success or failure. In any case, it is not my argument to recommend RJR. The argument is to show that compared with RJR, CJR is not really superior - not by much, anyway.
Before we proceed, an obvious objection needs to be answered. Surely, people would think, it matters what the constitution says. After all, there is a written constitution, with some determinate legal content, and it is the constitutional text and its legal content that judges need to implement by their decisions. Cases ought to be determined by the legal content of the constitutional text (and perhaps well- entrenched constitutional doctrines and precedents).
Therefore, the argument would go, the main difference between RJR and CJR consists in the fact that RJR is totally insensitive to the legal prescriptions embodied in the Constitution, whereas CJR is guided by the constitutional text, even if imperfectly so. Let me call this the obvious objection.It is difficult to answer the obvious objection in the abstract. The extent to which the content of constitutional documents actually guides constitutional decisions of courts varies a great deal between different jurisdictions. I will confine myself here to the U.S. model, and to the realities of constitutional judicial review in the United States.[6] So, here is the answer to the obvious objection: It is true that the constitutional text matters; the legal content expressed in the U.S. Constitution is not without significant legal ramifications. But the difference the constitutional text makes is rarely in play in the kind of cases that the U.S. Supreme Court decides on constitutional matters, for two main reasons: First, when the constitutional text evidently determines a given outcome, litigation is very unlikely to ensue. Parties have no money to waste on, and courts no patience and resources to deal with, cases in which a legal outcome simply follows from the public meaning of the relevant legislative text, be it constitutional or ordinary legislation. To put matters simply: Easy cases do not make it to the Supreme Court. If litigation makes it to the Supreme Court as a constitutional case, it is almost invariably because the text is not clear enough to dictate a particular result. I am talking about “the text” here, but we can easily extend the argument to include not only the text of the written Constitution but also deeply entrenched constitutional doctrines or precedents as well.[7] The point holds true even if we allow for a much broader sense of what constitutes “the constitutional text.”
Second, and this may be more unique to the U.S.
model, the Supreme Court itself gets a huge amount of discretion in determining the cases it is willing to hear. Only a small fraction of constitutional challenges filed get certified by the Court itself for hearing.[8] So the Court sets its own agenda, year by year, choosing from a wide variety of options. How does it make the choice? Obviously, the Court tends to choose the kind of cases in which it can make a difference. Naturally, those are the kind of cases in which reading the text and understanding what it says is just not going to suffice for a clear inference to the outcome. The Court would tend to grant cert in cases in which some reasonable argument can be made that the Constitution prescribes X rather than Y, or Y rather than X. In short, again, the Court would hardly ever grant cert to hear an “easy case,” one in which every competent lawyer would reach the same legal conclusion. That just does not happen.[9]To recap, briefly: The first assumption I make here, and one that I think is hardly controversial, is that if a constitutional case makes it to the Supreme Court it is not going to be the kind of case in which the constitutional text and deeply entrenched precedents, if you will, are simply going to determine a legal outcome. Constitutional cases at the Supreme Court level, at least, tend to be those in which plausible arguments can be made to interpret the Constitution one way or another, whereby none of the plausible readings is obviously dictated by the text. So there is that. And then, as we mentioned at the beginning, the result of the case is typically a function of the individual composition of the Court. Different justices would reach different conclusions, depending on their comprehensive moral, political and religious convictions. I am sure that one could give some exceptions and counterexamples. But I think we are entitled to assume here that, by and large, very few constitutional cases are actually determined, legally speaking, by the meaning of the text, by what the U.S.
Constitution simply says.[10] [11]The obvious objection may have a point, however, when you think about the legal impact of the constitutional text in those cases that do not make it to appellate courts, because the legal content of the Constitution is just clear enough to determine particular outcomes. In other words, supporters of CJR could claim that, even if my previous argument is correct, and easy constitutional cases do not make it to the Supreme Court, countless legal issues are determined by the Constitution simply because it is clear enough what the Constitution mandates or requires. That, of course, is quite true. The constitutional text, and probably even more so, the well- entrenched constitutional doctrines and precedents, make a significant legal difference in countless cases in which the legal content of the constitutional law is not in any serious doubt.
However, we can easily accommodate this concern by revising the hypothetical structure of RJR. Instead of assuming that all constitutional challenges are automatically submitted to the randomizing computer, we can confine the randomization mechanism to those cases that do make it to appellate courts, under the current CJR system, and fail to muster unanimous decision at the appellate level.11 The idea here involves a great simplification. It would take the unanimity of the decision by the appellate court as a proxy for cases in which the constitutional text, and perhaps deeply entrenched constitutional doctrines, are clear enough to determine particular results. And then, failure of unanimous consent on a constitutional case would be taken as an indication of some plausible controversy. Randomization would kick in, according to this revised system, only in cases of some actual legal controversy at the appellate courts level. Let us call the revised system RJR*. As I said, the use of unanimity at the appellate courts or the Supreme Court level should not be taken to be more than a simplifying assumption. It should be seen as a proxy for drawing the line between cases in which no serious legal doubt about the constitutional legal content can be raised, and those in which some plausible legal argument can be made to decide the case one way rather than another. It is not a perfect proxy, for sure, but good enough to make the argument here. Therefore, if you take the obvious objection to have a point, just think about RJR* instead of the original scheme; assume that randomization kicks in only in those cases in which there is some actual doubt about constitutional requirements. That would still cover the vast majority of cases that make it to the Supreme Court under the current CJR.
2.3