Concept
The Deferential Standard
Rational basis review
The Deferential Standard
The Equal Protection Clause forbids a State to “deny to any person within its jurisdiction the equal protection of the laws.”1 Earlier sections showed how a classification lands in the default tier: it burdens no fundamental right, it targets no suspect or quasi-suspect class, and, when the law is neutral on its face but falls unevenly, no one has proved a discriminatory purpose. What the Clause demands of such a law is the minimum test that every challenged classification must meet. The law stands if it is rationally related to a legitimate government purpose.2
That test asks two questions and then adds a procedural overlay. The first question is whether the law serves a legitimate purpose, an end that government may properly pursue. The second is whether the classification is rationally related to that end, so that the line the legislature drew could reasonably be thought to advance it.3 The overlay is that the challenger carries the burden of proof and a strong presumption favors the law.4 The overlay explains why the test forgives so much in practice: a court may sustain a law by pointing to any legitimate purpose that can be conceived for it, even one the legislature never had in mind.5
Courts review so leniently because they treat line-drawing among groups as work for legislatures. Nearly every statute favors some people over others, and those who dislike a given choice can seek its repeal at the polls. Since 1937 the Court has deferred to economic and social regulation unless it infringes a fundamental right or discriminates against a group that warrants special judicial protection, an approach defensible as judicial restraint that leaves decisions to the more democratic branches.6 Rational basis is accordingly the most deferential of the standards courts use in due process and equal protection analysis.7
Deference has a price. Unfair laws can stand, because a conceivable legitimate purpose can be found for virtually any law, and the Court seldom holds that a law fails the test.8 The decisions that strike a law while formally applying the lenient test are the exceptions.9 The three parts below take the elements of the standard one at a time, testing each against the police-retirement statute in Murgia and the classic economic-regulation cases.
Legitimate Government Purpose
The first question asks what end the classification serves, and Massachusetts Board of Retirement v. Murgia answers it cleanly. Massachusetts required uniformed state police officers to retire at fifty, and the Court identified the State’s purpose as protecting the public by assuring the physical preparedness of its uniformed police.10 The record supported that account. A special legislative commission’s report preceding the enactment described the work as arduous in the extreme and as requiring comparatively young men of vigorous physique.11 Murgia is the easy case because the end the State named, the end the record shows, and the end the Court accepted all coincide. The doctrine does not demand that coincidence.
Legitimate ends are easy to find. In Village of Belle Terre v. Boraas the Court sustained a village land-use ordinance and treated the creation of a quiet place, with wide yards, few people, and restricted motor vehicles, as a permissible goal, because the police power is ample enough to lay out zones where family values and the blessings of quiet seclusion make an area a sanctuary.12 Institutional security qualifies as well. In the prison-regulation cases taken up at the end of this chapter, officials testified that mail between institutions can carry escape plans and arrange assaults, and that restricting correspondence among gang members was an important element in combating prison gangs.13 Because a conceivable legitimate purpose can be identified for virtually any law, the first question seldom decides a case.14
A legitimate purpose also need not be the one the legislature held or announced. In FCC v. Beach Communications, Inc., the Court said it is “entirely irrelevant for constitutional purposes whether the conceived reason for the challenged distinction actually motivated the legislature,” because a legislature never has to articulate its reasons.15 The federal statute there required cable systems to obtain local franchises but exempted systems serving only commonly owned buildings. Nothing in the purpose of the franchise requirement obviously explained the exemption, yet the Court upheld it on two conceivable justifications: a system serving commonly owned buildings is likely to be small, so a franchise application would burden it more, and consumers can negotiate more easily with a single owner, so city approval is less necessary.16 U.S. Railroad Retirement Board v. Fritz took the same view of a line Congress drew among railroad retirees. The consequence is that government counsel may supply the purpose for the first time in litigation.
Two older decisions show how little evidence a hypothesized purpose needs. In Williamson v. Lee Optical of Oklahoma, Inc. Oklahoma barred opticians from duplicating or replacing lenses without a prescription from an ophthalmologist or optometrist, and the Court upheld the rule because the legislature might have thought that requiring an eye examination whenever glasses were replaced would improve eye care. In Railway Express Agency v. New York, the Agency sold the space on the sides of about 1,900 trucks in New York City, mostly for advertising unconnected with its own business, and was convicted.17 The city’s rule prohibited advertising on the sides of vehicles unless the ads promoted the vehicle owner’s business, and the Court upheld it, speculating that the legislature might have thought such ads less likely to distract drivers, although no evidence supported that guess.18 For the Court to uphold a law this way, it seems necessary only to imagine a hypothetical rationale.19 Since a rationale can be imagined for almost any law, laws enacted to help a politically organized group at the expense of others can ride the same latitude.20
The Court accepts a purpose that only counsel has supplied partly because the Constitution presumes that even improvident social and economic decisions will eventually be rectified by the democratic process.21 The cost of that trust is that a law written to favor a politically strong group can survive whenever some legitimate purpose can be imagined for it.22
The freedom to hypothesize belongs to rational basis review alone. When a law draws gender lines and intermediate scrutiny applies, the justification must be genuine, not hypothesized or invented after the fact in response to litigation, as the Court said in United States v. Virginia.23 A litigation-stage rationale that rescues a classification in the default tier therefore fails in a heightened tier. Identify the tier before deciding whether counsel’s explanation counts.
Counsel’s freedom narrows when the legislature has itself spoken. When state law identifies a specific purpose for a classification and leaves no room for other conceivable purposes, the stated purpose controls. Nordlinger v. Hahn upheld California’s property tax scheme and distinguished it from a nearly identical scheme invalidated in Allegheny Pittsburgh Coal Co. v. Webster County, because in the latter case the purpose identified by state law precluded reliance on other conceivable purposes.24 A classification may also be invalidated if it is clear that its actual purpose was illegitimate.25
The generosity stops at the character of the end itself. In United States Department of Agriculture v. Moreno, Congress amended the Food Stamp Act to deny aid to any household containing an individual unrelated to the other members, and the legislative history showed that the amendment was meant to keep so-called hippie communes out of the program. The Court held that a bare desire to harm a politically unpopular group cannot be a legitimate governmental interest, and that a classification clearly intended to injure a particular class, with only incidental or pretextual public justifications, must be struck down.26 Moreno is one of the decisions in which the Court has struck a law under rational basis review.27
Murgia and Moreno mark the two ends of the legitimacy question. Protecting the public is an objective the Clause tolerates, while disadvantaging a disfavored group is the harm the Clause exists to prevent, so a classification resting only on that aim has no legitimate purpose to which it could be rationally related. Cleburne and Romer v. Evans, taken up later in this chapter, are decisions in which the Court used rational basis review to invalidate a zoning ordinance that blocked a home for the mentally disabled and a Colorado voter initiative.28
An exam stem on this element typically supplies a classification with no recorded explanation and a government lawyer who offers a plausible one. The missing record does not help the challenger, because a conceivable legitimate purpose can be found for almost any law.29 The answer changes only if the statute itself recites a narrower purpose or the facts show that the sole aim was to injure a disfavored class, with only pretextual public justifications.30 Ask first whether government may pursue the end at all. A safety, health, or welfare goal passes that question easily.
Rational Means-Ends Fit
Once a court accepts the government’s end as legitimate, the second question is whether the classification rationally furthers it. Massachusetts Board of Retirement v. Murgia supplies the model. Because physical ability generally declines with age, the Court reasoned, retiring uniformed officers at fifty removes from the force those whose fitness has presumptively diminished, and that is rationally related to the State’s objective.32 The trial record backed the generalization. Physicians testified that the risk of physical failure, particularly in the cardiovascular system, rises with age and that the number of people in an age group who cannot perform stress functions grows as the group gets older, though they also recognized that particular individuals over fifty could perform safely.33 The word presumptively carries the logic. The State did not claim that every officer at fifty is unfit. It claimed that age is a workable stand-in for fitness, and a stand-in that tracks the trait in the aggregate is all the fit requirement asks. The Court also marked a floor: nothing indicated that the statute removed so few unqualified officers that age fifty had become a criterion wholly unrelated to its objective.34
The retired officer’s best argument was that Massachusetts already measured fitness directly. Uniformed officers had to pass a physical examination every two years until forty and every year after that, so the officers still serving at fifty had been repeatedly found fit.35 The District Court agreed and held the age line irrational in a scheme that assessed officers individually, because no one had shown that reaching fifty forecasts even imminent change in an officer’s condition.36 The Supreme Court reversed. The State’s choice not to test fitness individually after fifty does not mean the age limit fails to further the goal; it means only that the State perhaps has not chosen the best means, and a State does not violate the Equal Protection Clause merely because its classifications are imperfect.37 A footnote turned the existing examinations into support for the State. Annual testing between forty and fifty augments the scheme’s rationality, because it eliminates younger officers who are not at least as healthy as the best fifty-year-old, which carries out the legislative judgment that even the healthiest fifty-year-old presents a risk worth avoiding.38 A more precise alternative therefore counts for nothing against the State in this tier. The inquiry is whether the chosen means advance the end, not whether they are the best means available.
The tolerance follows from how classification works. A law is overinclusive when it reaches people the purpose does not require it to reach, such as the healthy officer retired on his birthday, and underinclusive when it leaves out others who share the problem. Every line leaves some individuals on the wrong side of it, and the Court treats the drawing of such lines as a legislative task for which perfection is neither possible nor necessary.39 Age lines show the point. A fourteen-year-old denied a driver’s license prevails only by proving that the law is not rationally related to a legitimate government purpose.40 A highway speed limit illustrates the same tolerance: seventy miles per hour is a sensible limit and ten or one hundred twenty is not, yet the choice among 68, 70, and 72 is necessarily arbitrary, and courts must accept that arbitrariness at the margin.41
Overinclusion draws the same tolerance outside age cases. In New York City Transit Authority v. Beazer, the Transit Authority refused to employ anyone in methadone treatment. The Court upheld the rule, reasoning that any special rule short of total exclusion would be less precise and costlier, and that eligibility after one year of treatment or at any later point would still divide employees equally or almost equally likely to recover.42 Even the District Court did not treat one year as a constitutionally mandated measure.43 The dissent in Craig v. Boren voiced the same attitude toward numbers: the rationality of a classification does not depend on the statistical “fit” between the class and the trait, and the question is whether the trait appears often enough more in the included class to justify different treatment.44 The Craig majority did not accept that approach for a gender line. Applying heightened scrutiny, it held that the State had not shown the classification substantially related to traffic safety.45 The dissent describes how loosely the lenient test measures fit. It does not describe how courts review gender lines.
Underinclusion is tolerated for a different reason: reform can proceed in stages. A statute challenged as underinclusive may be sustained because legislatures may take a piecemeal approach and deal with a general problem a step at a time.46 Lee Optical applied the idea when Oklahoma required a prescription before opticians could duplicate lenses but exempted sellers of ready-to-wear glasses, and Railway Express applied it when the city banned vehicle advertising for outsiders but let owners advertise their own business. A rule requiring a legislature to attack a whole category of harm or none would often leave the harm untouched. Practical enforcement can also justify a line. A Sunday closing law that lets amusement park vendors sell selected items but not department stores may stand, because a broader exemption would be difficult to enforce against the department stores.47 The cost of tolerating underinclusion is the one noted earlier: a law that favors a politically strong group can stand if its line has any conceivable connection to a legitimate end.48
The prison-regulation cases taken up at the end of this chapter show the same refusal to demand the best means. The test there is not a least-restrictive-alternative test, so officials need not set up and shoot down every conceivable alternative, although an alternative that fully accommodates the prisoners’ rights at de minimis cost may count as evidence that the regulation is unreasonable.49 The proposed alternative of monitoring inmate correspondence failed that test. Officials testified that it would be impossible to read every piece of mail, that dangerous messages would be missed, and that the burden on staff would be heavy.50
The tolerance has a floor, and the floor is the reason the fit element exists at all. Murgia itself asked whether age fifty had become a criterion wholly unrelated to the statute’s objective.51 Romer v. Evans applied the floor to a Colorado voter initiative that repealed laws prohibiting discrimination based on sexual orientation and precluded new protections, and the Court held that the initiative failed rational basis review.52 The reasoning is taken up later in this chapter.
Moreno, one of the laws the Court has struck under rational basis review,53 shows the floor from the other side. The Government defended the exclusion of households containing unrelated members as a fraud-prevention measure, a legitimate end. In practical operation, though, the rule excluded households too desperate to rearrange their living arrangements, not households likely to abuse the program, and the Act already contained separate fraud provisions. Classifications need not be drawn with mathematical nicety, but this one was wholly without a rational basis, which is different from imprecise. The Texas Supreme Court drew the line the same way in Sullivan v. University Interscholastic League, where a rule made every transferring non-senior ineligible for varsity football and basketball for a year in order to discourage recruiting. The court stressed that the rule was overinclusive in light of its purpose and rested on an irrebuttable presumption, and it held the rule not rationally related to that purpose.54 Murgia’s age rule was overinclusive too, so overinclusion alone cannot separate the cases. What separates them is the strength of the link: age tracks physical decline strongly enough that many of the officers retired were in fact unfit, while a rule aimed at recruiting that sweeps in many innocent transfers tracks its target poorly.
The second Sullivan factor needs care on a federal exam. Murgia’s age rule is itself an irrebuttable presumption that officers over fifty are unfit, and it was upheld. Vlandis v. Kline had condemned permanent irrebuttable presumptions that are not universally true when reasonable alternative means of making the determination exist.55 Weinberger v. Salfi then narrowed that doctrine to cases where heightened scrutiny would be called for, so that otherwise the government may legislate through generalizations if it can rationally conclude that generalized rules suit its purposes.56 The absence of an individualized hearing therefore does not defeat a law reviewed for rationality. The question remains whether the generalization has a rational link to the purpose.
Justice Marshall’s dissent in Murgia shows the fit element by contrast. He relied on the State’s concession that an officer who passes the annual physical is found qualified to perform all the duties of the uniformed branch,57 and he would have required the State to show a reasonably substantial interest and a scheme reasonably closely tailored to it.58 That argument has force only under a demand for close tailoring. Once the majority placed age in the default tier, the measure of fit became whether the age line plausibly tracks fitness in the aggregate, and the existence of fit individuals past the line did not answer that question.
An exam stem on this element usually features a plaintiff who is individually harmless or fully qualified and who is excluded anyway. That plaintiff describes overinclusion, and overinclusion alone does not win. Ask instead whether the trait the law uses tracks the problem it addresses. A massage ordinance that barred massages across the sexes in massage parlors, health salons, and physical culture studios, but not in barber shops, beauty parlors, or YMCA and YWCA health clubs, fell because its class was structured arbitrarily, without rational relation to an evil that could as easily occur in the unregulated places.59 The Sunday closing law, by contrast, survived its exemption because a broader exemption would have been hard to enforce.60 The first line ignored the problem, and the second followed it.
Presumption and Burden of Proof
Rational basis review starts from a presumption, and the presumption decides who loses when the record is thin. In Massachusetts Board of Retirement v. Murgia the Court called its standard relaxed, because drawing lines that create distinctions is peculiarly a legislative task, and it added that such legislative action is presumed valid.62 The preceding parts taught the two questions the default tier asks. This part asks who must prove what on each. The challenger carries the burden of proof, and a strong presumption favors the law.63 A party who bears a burden loses whenever the record leaves the issue unresolved, so that allocation shapes most rational basis cases before any evidence is weighed.
The burden has an unusual shape. A challenger cannot win by showing that the reason the government gave is weak, or even false. In FCC v. Beach Communications, Inc. the Court said that those attacking the rationality of a legislative classification have the burden “to negative every conceivable basis which might support it.”64 The challenger must therefore eliminate every justification a court can imagine, including justifications that no legislator mentioned and that the government’s lawyer first thought of in litigation. Proving a negative that broad is the real difficulty, and a claim against ordinary economic or social legislation usually fails. The allocation fits a general judicial reluctance to reject a legislature’s judgment outright and foreclose experimentation with other formulations.65
Because the court assumes the supporting facts, the government has no duty to offer any. Under the formulation traced to Lindsley v. Natural Carbonic Gas Co., if any state of facts reasonably can be conceived that would sustain a classification, the existence of those facts when the law was enacted must be assumed.66 A legislature need not make findings either. Even without them, the existence of facts supporting the legislative judgment is presumed, and the question under United States v. Carolene Products Co. is whether any state of facts, known or reasonably assumed, supports it.67
Central State University v. American Association of University Professors shows what that means for a challenger who points to an empty record. The professors observed that the State had produced no evidence that making teaching workloads a subject of collective bargaining would reduce classroom time. The Court answered that a government defending a classification subject only to the rationality test has no obligation to produce evidence, so long as a legislature could rationally conclude that bargaining over workloads might interfere with its goal of improving instruction.68 Beach Communications later said that a legislative choice is not subject to courtroom fact-finding and may rest on rational speculation unsupported by evidence or empirical data.69
The same allocation limits what a challenger’s own evidence can accomplish. In Houser v. State, a class of 18- to 20-year-olds attacked a minimum drinking age of 21 on equal protection grounds. Their expert’s affidavit asserted that several traditional arguments for the age limit lacked scientific support, while the State submitted technical studies supporting it. The court granted summary judgment to the State, explaining that the motion required it to inquire not into the facts of one dispute but into the general relationship between reaching age 21 and the effects of alcohol, a question essentially of law about whether the distinction was rationally related to the State’s purposes.70 The test asks what it was rational to believe, so a conflict between experts about whether the line actually works does not entitle the challenger to a trial. The State prevails if the belief was rational, whichever expert proves right.
Railway Express Agency v. New York shows the burden operating on its home ground. The Agency sold advertising space on the sides of about 1,900 trucks in New York City, mostly for businesses unconnected with its own, and was convicted under a rule that barred vehicle advertising unless the ads promoted the vehicle owner’s business.71 The Court sustained the rule by speculating that the legislature might have thought owners’ own ads less likely to distract drivers, although no evidence was presented.72 The city had to prove nothing, while the Agency had to show that no such conclusion was conceivable, and it could not.
The allocation reverses when the classification is racial, which shows how much of the rational basis result flows from the burden alone. In McLaughlin v. Florida, a statute punished unmarried interracial couples who habitually shared a room at night, and Florida defended it as a guard against breaches of sexual decency. The Court did not ask whether the statute had some conceivable rational basis. It placed the burden of justification on the State.73 In cases of this kind the statute is presumed unconstitutional, and the State can overcome the presumption only by establishing that the classification is necessary, and not merely rationally related, to a permissible state policy.74 Florida’s decency defense did not carry that burden. Loving v. Virginia drew the contrast directly. In cases not involving race the Court has merely asked whether any rational foundation exists, but statutes containing racial classifications face the very heavy burden of justification that the Fourteenth Amendment has traditionally required, and equal application does not immunize them.75 Under strict scrutiny generally, the government must justify the law as necessary to a compelling purpose, and a loose-fitting law is likely to fall. Both Railway Express and McLaughlin involved a classification defended by a conceivable public concern, traffic safety in one and decency in the other, so the justification does not separate them. The classification does, and with it the party who must come forward.
Each half of the allocation has its own reason. The presumption of validity rests on accountability: the Constitution presumes that even improvident social and economic decisions will eventually be rectified by the democratic process, so a court has less cause to step in.76 The burden on the challenger rests on the character of the question. Whether a classification furthers a social or economic goal is a matter of policy judgment, and the Court said in Railway Express that it does not sit to weigh evidence in order to decide whether a regulation is sound or appropriate, or to pass judgment on its wisdom.77 A rule requiring the government to prove its case would turn every challenge into a trial about the wisdom of legislation, and courts are reluctant to reject legislative judgments outright in that way.78 The reversal for racial classifications applies the same logic in reverse. Classifications based on alienage, like those based on nationality or race, are inherently suspect and subject to close judicial scrutiny, and aliens as a class are a prime example of a discrete and insular minority for whom heightened judicial solicitude is appropriate.79 Where the political process cannot be trusted to correct the law, the reason for presuming validity falls away and the State must justify the law.
Together these features bring review in this tier close to non-review. One authority has said the traditional standard “borders on being a rule of non-review.” In practice the choice of tier usually settles the outcome, because courts rarely find a compelling interest and usually find a rational basis.80 A burden placed on the government under a rational basis label is therefore a signal worth noticing. In City of Cleburne v. Cleburne Living Center, Inc., the Court concluded that persons with intellectual disabilities are not a suspect or quasi-suspect class and recited the rational basis standard. Yet it seemed to apply a form of heightened scrutiny, later dubbed “rational basis with bite,” by imposing the burden of proof on the party seeking to uphold the ordinance, contrary to traditional practice.81 A shifted burden suggests heightened scrutiny operating in disguise, a possibility taken up in the discussion of rational basis with bite later in this chapter.
On an exam, the tier answers the burden question, so the first move is to identify the classification. If the stem leaves the law in the default tier, a challenger who produces studies, expert testimony, or statistics suggesting that the law does not work has shifted nothing. The government’s failure to offer evidence or an actual legislative reason gives the challenger no help either, because the government’s lawyer may supply any conceivable rationale. The usual error is to fault the state for failing to prove that its classification advances its goal, which is the question strict scrutiny asks. A stem that makes the state show that the law is necessary, or that describes a court presuming the law invalid, has left the default tier.