Eugene's most recent post notes yet another case where courts have purportedly applied "rational basis" scrutiny to a law, but in reality have held it to a stricter standard. If judges believes that a particular classification deserves a less deferential form of scrutiny, they may well have good reason for reaching that conclusion. But instead of pretending to apply the rational basis test, they should explicitly indicate that they are applying a higher standard, such as "intermediate scrutiny" (the level of scrutiny now applied to sex-based classifications), or "strict scrutiny (applied to racial, ethnic, and religious classifications).
Traditionally, a law that is only subject to "rational basis" review by courts applying the Equal Protection Clause of the Fourteenth Amendment, would almost certainly be upheld. The law or regulation need only be "rationally related" to a "legitimate government interest," with no real judicial scrutiny of 1) the likelihood that the law actually will succeed in promoting the "interest" in question, or 2) the degree to which the enactment of the law actually was motivated by the supposed interest. The whole point of the rational basis test was to ensure that only most incredibly moronic of laws using a given classification would be struck down.
In recent years, however, courts have increasingly used the rational basis test to strike down laws that would almost certainly be upheld under the traditional ultradeferential version of the test. Eugene's post on the Tenth Circuit's decision to strike down Oklahoma's rule for considering the admission of nonpublic schools to the state athletic association is a noteworthy recent example; as Judge McConnell explains in his concurrence, the rule would almost certainly be upheld under the standard form of rational basis review. Perhaps the most famous example is <Romer v. Evans, a 1996 case where the Supreme Court essentially held that some kinds of laws discriminating against homosexuals, although officially subject to rational basis review, would in reality have to pass a more stringent version of the test than that applied to most other laws.
In many cases, I think that increasingly stringent judicial scrutiny is justified. But courts that decide to hold a given type of classification to a higher standard should be open about what they are doing, instead of pretending to apply the traditional rational basis test. Otherwise, it will be difficult to distinguish between those classifications that are still subject to the traditional highly deferential rational basis scrutiny, and those that must pass a stricter "rational basis plus" standard. This point is particularly relevant to appellate courts, especially the Supreme Court. Perhaps the most important function of appellate courts is to provide clear rules for lower courts to follow in making decisions.
To be sure, it may be that Supreme Court justices would like to continue to fudge the meaning of the rational basis test in order to increase their discretion in deciding future cases. A law discriminating against homosexuals that the justices dislike (such as the one in Romer) can be struck down by applying "rational basis plus," while one that they prefer to leave alone (such as a ban on homosexual marriage), can be upheld under the traditional rational basis test. And all of this can be done without ever explicitly committing the Court to any clear general rule governing judicial scrutiny of sexual orientation classifications. However convenient for the Supreme Court, this kind of sleight of hand undermines the transparency of the legal system, and also makes it difficult for lower courts to apply Supreme Court precedent consistently. If the Supreme Court majority believes that laws discriminating on the basis of sexual orientation require greater scrutiny than that imposed on ordinary legislation, they should write an opinion explicitly stating that holding, and outlining the higher standard that such classifications must meet.
It may be unoriginal to suggest that appellate court decisions should mean what they say and say what they mean. But it's a good - and even "rational" - rule nonetheless.