Bob Nagel is a well-known and persistent critic of Supreme Court decision making—and in particular, the Court’s stated formulae for how those decisions are reached.1See generally ROBERT F. NAGEL, CONSTITUTIONAL CULTURES: THE MENTALITY AND CONSEQUENCES OF JUDICIAL REVIEW (1989); ROBERT F. NAGEL, UNRESTRAINED: JUDICIAL EXCESS AND THE MIND OF THE AMERICAN LAWYER (2008). Bob’s neologism, “the formulaic Constitution,” was not coined to be an honorific term. For Bob, the Court’s announced rationales often seem hollow and thus quite manipulable, disguising whatever might have been the Court’s real reasons for reaching a decision and perhaps even blinding the Court itself to those reasons.
Those of us who are unfortunately tapped by our deans to teach constitutional law feel doomed to disappoint students who quite rightly expect a payoff of doctrines that are reasonably coherent. For in most areas of constitutional law, doctrine as such is absent and what exists in its place is one damn case after another, often with shifting five-to-four line- ups of Justices. Perhaps each justice is consistent—though I doubt it—but the nine-Justice ensemble, with each of the nine eventually replaced by someone with a different view of the cathedral, is surely not consistent. Thus, in most areas of constitutional law, a lawyer who cannot find Supreme Court decisions that support her side of a case ought to be disbarred for incompetence, as should a lawyer who cannot see that there are Supreme Court precedents supporting the other side. So instead of teaching law students a body of doctrine, we constitutional law teachers are forced to teach “one can argue this and cite that,” and “one can argue the opposite and cite that.” Law is a normative discipline, but that is about as normative as we constitutional law teachers can get and still tell the truth. We could, of course, discuss the original meaning of constitutional provisions, which might be useful if the Court paid any attention to the Constitution’s meaning. Or we could mimic the social scientists and try to be predictive, which is useful in advising clients perhaps, but of no use when you must make normative arguments to a court.
Bob saw early on—just how early I’ll get to momentarily— that the Court’s constitutional standards of review were quite arid and manipulable formulae that did no work and could do no work in deciding cases. Instead, those standards disguised whatever actual grounds for decisions were consequential. Has the Court ever really told us what governmental purposes, described at what level of abstractness, are “legitimate,” or “important,” or “compelling”? Indeed, has it ever told us from where in the Constitution or its context such a hierarchy of governmental purpose is derived? Those questions are, of course, rhetorical. Their answers are “no” and “no.”
As I said, Bob saw all this early on—as a law student, no less. When I began teaching law in 1970, I had a subscription to the Yale Law Journal, perhaps because I was a Yale law product, but more likely because I imagined a few prestigious law journals would be sources of much-needed wisdom—a view from which I was eventually weaned. In any event, as I was teaching constitutional law, I read all the constitutional law articles in those journals, including the student notes. And although the articles seldom provided the wisdom I sought, one student note did. That note was titled Legislative Purpose, Rationality, and Equal Protection,282 YALE L.J. 123 (1972). and its author was one Robert Nagel, Yale law student.
Bob’s note, which I recommend to all of you, is brilliant. I would place it alongside Anthony Amsterdam’s justly famous note on the void-for-vagueness doctrine as maybe the best student note ever published. (Indeed, I think Bob’s note is even better than Amsterdam’s.)
Bob’s note targets the Court’s rational basis review. The cases he analyzed may be unfamiliar to many of you. After all, the note was published in 1972, almost two generations ago. Still, Bob gives you all the facts and opinion passages you need to get his point. And that point is that rational basis review cannot work as the Court claims it does.
In a nutshell, when a statute’s non-suspect or non-semi- suspect classification is challenged as violative of equal protection, the statute will be upheld if, but only if, the classification is “rationally related” to the legitimate purpose the classification is meant to serve. That is rational basis review, or so the Court says.
Bob begins his analysis of rational basis review by taking up the Court’s then-recent decision in Eisenstadt v. Baird,3405 U.S. 438 (1972). a case that I suspect most of you who teach constitutional law are familiar with, no matter your age. Recall that in that case the Court struck down a Massachusetts statute that made it a crime to dispense contraceptives to unmarried persons for the purpose of preventing pregnancy. The statute permitted married persons to get contraceptives for that purpose, so the equal protection question was whether distinguishing unmarried from married persons in that context served a legitimate state purpose. And the Court’s answer was that it did not.
As Bob pointed out, the Court reached this result by a “divide and conquer” method.4Nagel, supra note 2, at 127. It took up a series of possible purposes one by one and purported to show that none could support the differential treatment of the married and unmarried. The Court first took up the purpose of discouraging pre-marital sex, but found that because unmarried persons could get contraceptives to prevent disease, and married persons could use contraceptives to have sex with the unmarried, the prohibition of unmarried persons getting contraceptives to prevent pregnancy would have only a minimal effect on achieving its goal.
The Court then took up a second possible purpose, that of regulating the distribution of potentially harmful contraceptives by requiring them to be prescribed by physicians. But as the Court noted, the protective function of physicians applies equally to the unmarried and the married and so cannot support the statute’s differential treatment.
The third possible purpose was that of preventing the use of contraceptives. But because the statute did not prevent married people from using contraceptives, that purpose was obviously not served.
Here is Bob’s critique of the Eisenstadt Court’s rational basis analysis:
The “divide and conquer” analytical tactic also permitted a convenient simplification of each purpose, for it is only in the context of the full statutory scheme that full meaning can be given to each legislative objective. For example, when the plurality opinion discussed the third objective (preventing the general use of contraceptives), it asserted that evil “as perceived by the State” was identical as between the use of contraceptives by married and by unmarried people. The plurality could then argue that exemption of married persons from the prohibition against obtaining a prescription was not rational, for if contraception itself is evil, is it not as evil for married as for unmarried persons? But, if the legislation was also designed to discourage premarital sexual activity, the use of contraceptives by unmarried persons manifestly is not the same evil as is their use by married persons.
These illustrations show that if the plurality had defined the overall legislative purpose as consisting of the partial achievement of several sub-purposes, the determination that the statute was not rational would not have been so easy. The legislature’s overall purpose might have been defined as follows: to discourage premarital sex by making contraceptives harder to obtain to the extent that this would not increase the risks of venereal disease; to provide for the medical supervision of the distribution of contraceptives to the extent that this would not increase the availability of contraceptives to the unmarried; and to discourage the use of contraceptives to the extent that this would not interfere with the private behavior of married persons. Unless it is “irrational” per se for legislature to design a statute to achieve a set of somewhat conflicting policy objectives, the Massachusetts statute would appear to have been rational.5Id. at 127–28.
Bob argues convincingly that rational basis review is, as he puts it, “invariably an empty requirement and a misleading analytic device.”6Id. at 128. For, he points out, “[i]t is always possible to define the legislative purpose of a statute in such a way that the statutory classification is rationally related to it.”7Id. Because when a class is benefitted or burdened by a statute, the statute’s purpose can be deemed to be to benefit or burden the trait that defines the class.
Indeed, just as the Court in Eisenstadt disaggregated a possible mix of purposes and found each one wanting, when it wishes to uphold a statute it can aggregate distinct purposes and find that, voila!, the statute serves that mix perfectly. A couple of cases illustrate this method. Magoun v. Illinois Trust & Savings Bank8170 U.S. 283 (1898). dealt with an inheritance tax. The rate of the tax varied among blood relatives according to how closely they were related to the decedent—the closer, the lower the rate. And the rate varied among unrelated legatees based on the amounts of their bequests. The purpose to be derived from this scheme Bob states as follows: “The purpose is to reduce transfer of wealth through inheritance but to do so to the extent compatible with encouraging intra-familial support.”9Nagel, supra note 2, at 130. And the statutory scheme fits this purpose—which is really a mix of purposes—like a glove. Indeed, it fits it tautologically, for the purpose or mix of purposes is derived from the statutory classifications themselves.
Bob illustrates this tautological use of the rational basis test by comparing the majority and dissenting opinions in Levy v. Louisiana.10391 U.S. 68 (1968). The statute in Levy gave legitimate but not illegitimate children the right to sue for the wrongful death of their mother. Justice Douglas, writing for the majority, looked at only one purpose, that of compensating children for the loss of their mother, and could see no reason to distinguish illegitimate from legitimate children for that purpose. But Justice Harlan in dissent noted another purpose, that of discouraging non-marital procreative relationships. That suggests that the purpose of the statute could be construed to be compensating children for the loss of their mother to the extent compensation does not legitimize non-marital relationships.
One more case that is illustrative is another fairly old but probably quite familiar one, Railway Express Agency (REA), Inc. v. New York.11336 U.S. 106 (1949). There, New York City had a traffic regulation that banned advertisements on the sides of motor vehicles but exempted ads for the business of the vehicle’s owners. REA, which made a lot of money from carrying ads, usually for cigarette brands, on the sides of its fleet of trucks, claimed the exemption of owner ads violated equal protection. If the purpose of the ban of REA’s ads was to promote traffic safety by reducing driver and pedestrian distractions, then were not the owner ads equally distracting and equally disserving of the purpose? The Court surmised that the legislature had concluded that owner ads were less distracting, but that seems quite far-fetched. (Has anyone observed pest control vans recently, with their large rat or termite figures on the roofs?) The more accurate statement of the purpose, according to Bob, is to promote traffic safety slightly by reducing the number of distractions without jeopardizing the businesses who advertise on the sides of their own trucks12Nagel, supra note 2, at 144.— and, I would add, to do so without jeopardizing those merchants who have fixed signs at their places of business or jeopardizing the billboard industry and those who advertise on billboards. Stated thusly, the statute served its mix of purposes quite well.
Although Bob doesn’t mention it in his note, another purpose that when added to the mix can make classification schemes coherent is that of administrative convenience. Blunt rules, although frequently over- or underinclusive with respect to their primary purposes, are much easier to administer than more flexible standards or more complex rules. And ease of administration is, other things being equal, a good thing. For example, in Beazer v. New York City Transit Authority,13440 U.S. 568 (1979). a case decided many years after Bob’s note, the New York City Transit Authority banned methadone users from all jobs. The primary purpose was transportation safety, which former addicts might imperil if given certain jobs. But many jobs in the Authority were not of that type. Still, the Court upheld the ban for all jobs, in part because the blunt rule was easier to administer.
One way to avoid an empty tautological rational basis test is to deem some purposes per se illegitimate. The Court, however, has never told us what makes a purpose illegitimate other than a desire to harm a specific group. (Indeed, even that desire cannot always be illegitimate; can the legislature not desire to harm murderers and rapists?) Are there any other illegitimate purposes, and, if so, why are they illegitimate? And just where in the Constitution can one find the basis for calling some purposes legitimate or illegitimate?
Aside from deeming a purpose illegitimate, what other methods does the Court employ to avoid deeming classifications rationally related? Bob lists three: ignoring a purpose, stating the purpose as singular rather than a mix, and manipulating the level of abstraction at which the purpose is stated.14Nagel, supra note 2, at 132. Bob gives several examples of the first two, including the cases previously mentioned. So let us look at the third, the manipulation of the level of abstraction.
In Gulf, Colorado & Santa Fe Railroad Co. v. Ellis,15165 U.S. 150 (1897). a statute permitted successful tort claimants to recover attorneys’ fees if they had sued a railroad corporation. The Court held that the statute failed the rational basis test, but did so by focusing on purposes characterized sufficiently generally that the statutory distinctions failed to serve those purposes. As Bob put it:
The evils the statute might have been aimed at can be stated in the following order of descending generality:
(1) to decrease the likelihood that any tortfeasor will escape his obligation to repay his victim,
(2) to decrease the likelihood that any corporation might abuse the privilege of existence afforded it by the state by escaping its obligation to repay tort victims,
(3) to decrease the likelihood that any business corporation will abuse the privilege of carrying on business activities that the state affords it by escaping its obligation to repay tort victims,
(4) to decrease the likelihood that any quasi-public, commercial corporation will abuse its special function by escaping its obligation to repay tort victims,
(5) to decrease the likelihood that railroad corporations, which have a tendency to resist the demands of tort claimants, will escape their obligation to repay their victims.
The Court’s conclusion that the statutory classification was underinclusive followed inevitably from the level of generality at which it defined the evil at which the statute was thought to be aimed. The Court argued that the statute might have been intended to deal with problems as broad as (1) and as narrow as (4). Of course, if the Court had acknowledged the fifth purpose, a conclusion that the classification was rationally related to the purpose would have followed. The terms of the statute plainly suggest that the Court had defined statutory purposes too broadly.16Nagel, supra note 2, at 137–38.
Although there is much more in Bob’s note than I have discussed, I think I have adequately conveyed the strength, indeed the brilliance, of its critique of the rational basis test. It is hard to believe that Bob was only a law student at the time, but it is easy to see from Bob’s later work that the boy was the father of the man.