Cross-posted from Justia's Verdict.

Display of the Confederate battle flag on government-owned property has certainly gotten a lot of attention of late. The Supreme Court ruled 5-4 a few weeks ago, in Walker v. Texas Division, Sons of Confederate Veterans, Inc., that Texas could, consistent with the First Amendment, reject a specialty license plate design submitted by a non-profit entity due to the design’s prominent use of the Confederate battle flag, even though Texas had approved a disparate array of some 350 other designs submitted by various individuals and organizations depicting messages ranging from “I’d Rather Be Golfing” to the names of a number of (out-of-state) universities to “MIGHTY FINE” to “GET IT SOLD WITH RE/MAX.”

And, in the wake of the Charleston church shootings that occurred a day before Walker was handed down, there have been calls made and steps taken by (Republican as well as Democrat) elected leaders to remove the Confederate battle symbol from statehouses in places like Alabama and South Carolina.

Interestingly, the latter episode sheds light on how the Court should have handled the former—the Walker dispute. In short, the Court should have treated Walker more as a Confederate battle flag case, and less as a license plate case.

Why Justice Breyer’s Majority Opinion In Walker Is Unconvincing

Many able analysts, including some of my fellow Verdict columnists writing in other venues, have exposed some of the major shortcomings of Justice Breyer’s majority opinion, which holds that specialty license plates in Texas are “government speech,” such that the government is largely free to pick and choose—even on the basis of viewpoint—which messages shall be allowed on specialty plates mounted on Texas-registered vehicles. Justice Breyer’s opinion gives a few reasons why the Court believes specialty plates should be treated as government speech. First is that states have historically used license plates as a means of promoting government messages—such as a state’s motto or some iconic image of the state which its leaders want viewers to think of when the state comes to mind. Fair enough, but Justice Breyer is relying here on a prior era, and one in which Texas (and other states) had but a single, or a few, messages it allowed on any of the plates registered in the State. By opening up its license plate regime and approving over 350 disparate messages, Texas has changed the essential purpose (beyond identification of the car’s owner) of the license plate platform from government speech to private speech. And this is not a phenomenon unique to license plates. There might be a lot of public properties that once were used substantially for government speech, but that have been converted into limited or designated or nonpublic fora (where the government is no longer free to discriminate against particular viewpoints) by the government’s decision to open up the property so widely. And it is no response to a claim of improper viewpoint discrimination in any of these kinds of fora to say that the property was originally used for government speech.

The second argument Justice Breyer makes is that license plates are closely associated or identified with the state in the public’s mind. Justice Breyer adds that “a person who displays a message on a Texas license plate likely intend[s] to convey to the public that the State had endorsed the message. If not, the individual could simply display . . . a bumper sticker.” Again, that might have been true of Texas license plates in an early era, but as Justice Alito’s withering dissent points out, no remotely reasonable viewers would ever actually think, after driving Texas roads for even a short time, that Texas really supports messages like “Go [Michigan State] Spartans,” or “Roll [Alabama] Tide Roll.” Or that the person whose plate bears one of these messages actually chose the plate design over a bumper sticker in order to obtain a state imprimatur (as opposed to preferring a specialty plate over a sticker because of the permanence and cleanliness of a plate, and to avoid any stigma tied to the use of bumper stickers.) I also find it interesting that questions such as these are empirical ones, and yet the Court makes broad findings in this regard without any (easily obtainable) evidence. (This lack of empirical input concerning public perceptions was also a feature of another First Amendment case, Williams-Yulee v. Florida Bar, about which I wrote a column six weeks ago).

A Better Framing of the Issues Implicated by Walker

In spite of its analytic weaknesses, Justice Breyer’s opinion arrives at a result that many observers, on and off the Court, find intuitively correct. Can the Lone Star State really be forced to have the word “TEXAS” that runs accross the top of every license plate in the State appear just a few inches from the Confederate battle flag symbol? Walker, it is worth noting, was a rare case in which Justice Clarence Thomas joined the four “liberal” Justices to create a very unlikely 5-4 majority. Odd lineups like this usually suggest that something interesting is going on, but what might be the key to unlocking Walker?

For me, the critical fact in the case is one to which the Court gave virtually no attention—that the symbol at issue is not just an offensive one (representing, as it does, slavery, rebellion, and a combination of the two, whether or not it stands for other things as well.) It is a symbol that originated as government speech by, among other Confederate states, the State of Texas, and that is thus linked to those state governments, including Texas, who are responsible for its prominence and its (ugly) messages. The flag was adopted and used in battle by Confederate states from late 1861 until the fall of the Confederacy, and then utilized extensively by racist governors and other officials to resist desegregation during the Civil Rights movement. So even if Texas were required to permit racial slurs on specialty plates (a matter about which I’m not sure – perhaps Texas, because of its history concerning race, has a strong interest in separating itself from any racist message that, if attributed to the state, might raise equal protection problems), the present case is different: the Confederate battle flag is worse in this regard than a racial slur; it is akin to a racial slur that was adopted and previously used as a state’s motto.

Because Texas has a distinctive association with use of the Confederate flag, it has a distinctive (and non-censorial) reason to want to avoid improper attribution on its property, including its license plates. I am not saying that government owns the intellectual property rights to control the use of flag symbols—the famous Texas v. Johnson case invalidating a law prohibiting flag-burning rightly rejected that idea. Rather, because the risk of misattribution is greater with respect to the use on government property of symbols that had formerly been used by the government itself than it is with respect to other messages, the government has a correspondingly stronger reason (other than censorship) to regulate.

For example, if New Hampshire changed its motto from the traditional “Live Free or Die” message that has been appearing on its license plates for decades (and that was the subject of the other famous Supreme Court case involving a license plate, Wooley v. Maynard) because the State no longer embraced that principle, and it had a specialty plate scheme in which an applicant wanted that creed to physically appear near the words “NEW HAMPSHIRE,” the State would have a strong interest in rejecting that design to avoid any confusion.

If the approach I offer here had been the basis of the majority’s decision in Walker, various problems that Justice Breyer’s opinion creates going forward would have been avoided. If Texas specialty plates really are “government speech,” then how can the State approve and allow the design (which it has) bearing the words “Knights of Columbus,” a distinctively Catholic fraternal service organization? If people identify that design with lawmakers in Austin, isn’t there an Establishment Clause problem? And if the license plates are government (rather than private) speech, then Texas can reject a design bearing the words “Pro-Choice” even though it has already approved one bearing the words “Choose Life.” Does that kind of asymmetry make sense?

It may well be that the rationale I suggest today could, if adopted, mean that some states (i.e., those particularly associated with the Confederacy or racism) would have more latitude than others to reject the Confederate flag on specialty plate designs, and that the rejection by some states (that lack the historical association) would look like censorship, pure and simple. But even if that is true, we should remember that other equality-based constitutional doctrines are similarly contextual. For example, a jurisdiction that has engaged in racial discrimination may have more latitude to engage in race-based remedial action than one that hasn’t. The “government speech” rationale adopted by the Court may seem to create a cleaner line than the test I offer here, but in the long run it is (I suspect) less honest and will be more problematic in future disputes.

Cross-posted from Justia's Verdict.

On Monday, the Supreme Court handed down two cases, Elonis v. United States and EEOC v. Abercrombie & Fitch, that had received a lot of press in earlier stages of the litigations and that, judging from the briefs, posed important civil rights issues concerning freedom of speech and freedom of religion, respectively. Although the Court ended up resolving the two matters on relatively narrow grounds-disappointing some of the Justices as well as analysts-it is important to understand precisely what the Court did (and did not) hold in these two rulings, both of whose outcomes were decided by 8-1 votes. In the space below, I briefly discuss the two decisions individually and then side-by-side.

Elonis v. United States

Based on statements he posted on Facebook directed at, among others, his ex-wife, federal law enforcement officials, and school children, Anthony Elonis was convicted of violating federal criminal statutes that prohibit the interstate transmission of communications containing threats to injure other persons. For example, in referring to FBI officials (who had visited his home to interview him about his activities), Elonis wrote online (seemingly in rap-style cadence):

[T]he next time you know, you best be serving a warrant
And bring yo' SWAT an explosives expert while you're at it
Cause little did y'all know, I was strapped wit' a bomb . . .
I was jus' waitin' for y'all to handcuff me and pat me down.
Touch the detonator in my pocket and we're all goin' BOOM!

In another posting, Elonis offered:

That's it. I've had about enough.
I'm checking out and making a name for myself.
Enough elementary schools in a ten mile radius to initiate the most heinous shooting ever imagined. . .
The only question is. . . which one?

In entries about his wife, Elonis wrote: "There's one way to love you but a thousand ways to kill you. I'm not going to rest until your body is a mess, soaked in blood and dying from all the little cuts. Hurry up and die, bitch . . . " And so forth.

At Elonis's criminal trial, the federal district judge instructed the jury that, for purposes of whether Elonis had issued threats prohibited by the statute, "[a] statement is a true threat [subject to prosecution] when a defendant intentionally makes a statement in a context or under such circumstances wherein a reasonable person would foresee that the statement would be interpreted by those to whom the maker communicates the statement as a serious expression of an intention to inflict bodily injury or take the life of an individual." Elonis argued under both the First Amendment and also under the federal criminal statute he was charged with violating that, before a person can be punished for expressing a threat, the government must allege and prove that the defendant subjectively intended to threaten his victim. In other words, Elonis argued that the government needed to prove that he had some subjective state of mind with respect to the effect that his words would have on the individuals to whom they were directed, and not simply that he voluntarily uttered the words and should have known the effect those words would create.

The U.S. Court of Appeals for Third Circuit (along with a large number of other circuits) have rejected the idea that the First Amendment requires the government to prove any subjective state of mind in order to punish threats consistent with the First Amendment. Instead, the Third Circuit held that statements that are reasonably construed as threats by listeners can lawfully be punished. In contrast, the Ninth Circuit (and a number of state high courts) has read the First Amendment as requiring the proof of a subjective intent to threaten as a predicate to a prosecution for threatening speech.

When the Supreme Court granted review, most commentators expected it to weigh in on and resolve this divergence in the lower courts over what the First Amendment requires. But in its ruling three days ago, the Justices decided the case purely on the basis of the federal statute under which Elonis was prosecuted; the Court explicitly deferred any analysis or interpretation of First Amendment requirements. And under the federal statute at issue, the Court said, Elonis is correct that some subjective intent by the person uttering the alleged threat is required; negligence by the person issuing the threat (in the sense that he reasonably should have foreseen that his words would be interpreted as threatening) was not enough. Although Chief Justice Roberts's majority opinion conceded that there is no intent standard written into the text of the federal statute, the Court nonetheless found one based on the way similar statutes had been construed. The Court did not specify precisely what level of intent the federal government must prove-and explicitly left open the question whether recklessness (a conscious disregard of a known risk that words could cause fear) is enough for the government to prove, or instead whether a higher form of intent such as actual knowledge is needed-but reversed Elonis's conviction and sent the case back to the lower courts because negligence on his part was not adequate to support a conviction under the statute.

EEOC v. Abercrombie & Fitch

Samantha Elauf is a practicing Muslim who regularly wears a headscarf for religious reasons. When she applied for a job at an Abercrombie & Fitch (A&F) retail store, the A&F managers evaluating her candidacy declined to hire her because the wearing of head scarves violates an appearance (or "Look") policy A&F has; employees in retail positions are prohibited from wearing caps and other headwear. Prior to their decision not to hire Ms. Elauf, the A&F managers had internal discussions about her in which one of the managers who had some acquaintance with Ms. Elauf expressed the belief that Ms. Elauf wore headscarves because of her faith.

The Equal Employment Opportunity Commission sued A&F on Ms. Elauf's behalf, alleging that A&F had violated Title VII, which prohibits an employer from deciding not to hire an individual because of the individual's religious observance or practice, unless the accommodating the observance or practice would create an undue hardship for the employer. A&F argued, and the Court of Appeals for the Tenth Circuit agreed, that A&F should not be liable because "ordinarily an employer cannot be liable under Title VII for failing to accommodate a religious practice until the applicant . . . provides the employer with actual knowledge of his need for an accommodation."

The Supreme Court reversed. Justice Scalia, calling this a "really easy" case when he delivered the decision from the bench, authored the majority opinion which said a Title VII plaintiff need not show "actual knowledge [by the employer] of a conflict between an applicant's religious practice and a work rule," because requiring such actual knowledge would involve the Court "add[ing] words to the law to produce what is thought to be a desirable result [. . . , which] is Congress's province." He went on: "We construe Title VII's silence [as to the requirement of actual knowledge] as exactly that: silence."

But Justice Scalia did say that a Title VII plaintiff like Ms. Elauf had to show the employer's adverse action against the applicant was based at least in significant measure on a motive to avoid the religious accommodation. But how can an employer be acting based on a motive to avoid accommodating an applicant's religion practice-"because of" an applicant's religious observances, in the language of Title VII-if the employer doesn't know that the applicant's conduct requiring an accommodation is itself religiously based? Justice Scalia acknowledged, in an important footnote, that it "is arguable that the motive requirement itself is not met unless the employer at least suspects that the practice in question is a religious practice," but that issue need not be decided in the A&F case because it has not been briefed by either side and because A&F "knew or at least suspected that the scarf was worn for religious purposes." Oddly enough, then, the majority opinion technically holds open the possibility that an employer who had no clue that a prospective employee's likely non-compliance with a work rule was religiously motivated could be held liable under Title VII, a result that Justice Alito in his separate concurrence rightly calls "very strange" and "surely wrong."

The Two Cases Taken Together

What can we say about these cases when we look at them in tandem? First, it is interesting that Justice Clarence Thomas is the lone dissenter in both decisions, and that he parted company with Justices Scalia and Alito (with whom he is often aligned as to result) on the interpretation of the federal threats statute and Title VII. Also, Justice Alito wrote separately for himself in both cases, decrying the minimalist nature of the majority's holdings. So, in both Elonis and Abercrombie & Fitch, Justice Scalia, Justice Thomas, and Justice Alito each had a different take from one another.

Second, both majority opinions conclude that Congress had not fleshed out in the text of the statute in question a requirement as to the defendant's subjective knowledge, but the Court reacted to that absence differently in the two cases. In Elonis, the Court said that because the statute is a criminal one, a wrongful state of mind on the part of the defendant should be found in the statute even when Congress was silent. But in Abercrombie & Fitch, Justice Scalia's majority opinion insists (to the extent that one can separate motive from knowledge) that Congress's silence as to the level of knowledge required of employers before they can be held liable is not something to be fixed by courts. The more ambitious attitude by the Court in the context of a criminal statute makes sense; there are special rules of statutory construction that apply particularly to laws that impose criminal sanctions.

But, and this is a third point, the criminal nature of the statute in Elonis might have properly led the Court to want to provide more notice to lower courts and potential defendants about precisely what level of subjective knowledge concerning the fear-inducing nature of words is required; the criminal law setting usually calls for clearer notice to be given to potential offenders. In particular, the disinclination by the Elonis majority to weigh in on whether recklessness by a defendant-e.g., a defendant who fleetingly wonders whether his proposed speech might cause fear in others but who never forms a view on the likelihood fear will in fact ensue-is sufficient under the statute, while perhaps understandable given absence of explicit lower court consideration of this matter, is sub-optimal. And, of course, because many statutes that criminalize threats are state statutes instead of federal laws, the question of whether and how the First Amendment requires government to prove any particular mental state of a defendant before criminal punishment may be imposed remains one on which the Supreme Court will need to give guidance. Indeed, in an earlier column I wrote previewing Elonis, I noted that it may not have been a good case in which to grant review precisely because the statutory ground could complicate the ability of the Court to give needed constitutional guidance. If this happened (as it did), I observed, the Court "would still need to rule in a later case on whether the First Amendment requires subjective intent (in the context of a federal or state statute that clearly does not require it.)"

The absence of guidance to lower courts and litigants in Abercrombie & Fitch is also quite frustrating, even though criminal liability was not at issue. On the facts of the Abercrombie & Fitch case, the employer "knew-or at least suspected-that the scarf was worn for religious purposes." But the tougher situation arises when the employer doesn't actually know but perhaps should be encouraged to find out, prior to declining to hire someone. For example, what if the A&F managers had no personal knowledge of Ms. Elauf during the interview process, but noted that she wore a head scarf and said to themselves, "Gee, I wonder if that is a style statement or a religious practice?" (In some respects, that could be thought of as "recklessness" but not knowledge on the part of the employer.)

Would the fact that the possibility of religious motivation even occurred to the employer be enough to trigger a requirement that the employer investigate the basis of the practice? Or would imposing such a duty on employers cause them to invade the religious privacy of employees and job applicants? (Courts in other countries that take religious liberty seriously often focus on religious privacy more than do American courts.) I do not know the answer to this, but I would observe that minority religious practices are often less well-known to many employers, a fact which might argue in favor of requiring employers to do some diligence once the possibility of a religiously inspired practice occurs to an employer. As for respecting privacy, there may be sensible ways to avoid making applicants feel uncomfortable. For example, all prospective employees could be given a list of all the employer's work rules and then asked, as a matter of course, whether religious practices would require accommodations with respect to any such rules. But this precisely the kind of detail the majority in Abercrombie & Fitch did not want to wade into. The reason it was a "really easy" case for the Court is that the Justices shied away from the difficult matters that actually needed some clarification.

Cross-posted from Justia's Verdict.

A few weeks ago the Supreme Court handed down an important yet under-noticed case, Williams-Yulee v. Florida Bar, in which a 5-4 majority upheld a Florida law that forbids candidates running in contested elections for judicial office from personally soliciting campaign contributions, even though the state permits such candidates to raise money through surrogates (campaign committees) and also allows candidates to find out who contributed to their campaigns. In the space below, I identify four key takeaways from this "sleeper" ruling by the Court, a ruling that affords important insights about constitutional doctrine and also about the membership of the Roberts Court.

1. The Speech Clause Juggernaut May Be Losing Steam

The (unsuccessful) challenge to the Florida law was brought under the First Amendment; the defendant in Williams-Yulee argued that Florida's ban on personal solicitation was a regulation that singled out certain speech-a personal request for money-because of its content, in violation of free speech principles. The Court acknowledged that the Florida law was a content-based regulation of political speech (and, as explained in more detail below, thus purported to apply "strict scrutiny" to the matter), but nonetheless upheld the law because of the important countervailing interest in preserving public confidence in the integrity of the judiciary.

In holding that public perceptions of integrity should carry the day, the Williams-Yulee ruling stands in contrast to the great majority of free speech cases decided by the Court over the last generation. Since the early 1990s, the overwhelming majority of plausible free speech claims (and the defendant's claim in Williams-Yulee was certainly plausible) that have reached the Court have prevailed, and expressive autonomy has regularly trumped competing constitutional and societal values. Over the last quarter-century, the Court has invoked the Speech Clause to invalidate federal, state, or local laws and regulations in well over fifty cases, averaging close to three cases each year, a substantial number given the Court's small yearly docket of between seventy and eighty cases for most of that period.

But a quantitative inquiry tells only part of the story. It is particularly noteworthy that First Amendment claims grounded in expressive autonomy rights have not just been winning, but have been winning against-and requiring significant sacrifices of-other values that traditionally have enjoyed high esteem in our legal, social, and constitutional traditions, including the efficient functioning of labor unions, the protection of military honor and military families, antidiscrimination laws and norms, election and campaign finance regulation intended to make elections more free and fair, parental control over the upbringing of their children, and consumer protections, among others.

Whether Williams-Yulee represents simply one exception to this great tide of free speech victories, or instead should be viewed as part of the beginning of a more balanced approach to free speech cases remains to be seen. There are at least two (and maybe more) other interesting and difficult free speech decisions yet to be decided this Term. The first is a case that considers the extent to which the First Amendment protects against prosecution individuals who utter words that cause objectively reasonable people to feel fear (Elonis v. U.S.), and the second is a case about how readily a State can discriminate among messages on personalized automobile license plates (Walker v. Texas Division, Sons of Confederate Veterans). It is possible that the free speech claimants in both of those cases (who assert plausible, if to my mind flawed, free speech arguments) will also lose. If that happens, commentators will begin to wonder whether the free speech juggernaut is indeed beginning to slow.

2. "Strict Scrutiny" Is in the Eye of the Applier

As I noted above, the Court in Williams-Yulee applied strict scrutiny-which requires the government to prove that the law in question is narrowly tailored to serve a compelling interest-to the Florida election regulation. But, as Justice Scalia remarked in dissent, "[although the Court] purports to reach [its] destination by applying strict scrutiny, . . . it would be more accurate to say that it does so by applying the appearance of strict scrutiny." In particular, the Court seemed quite tolerant of underinclusiveness in Florida's scheme, whereas significant underinclusiveness usually prevents a statutory scheme from being considered "narrowly tailored" in the way that strict scrutiny dictates.

For example, the defendant pointed out that Florida permits candidates to write personal thank-you notes to donors (guaranteeing that the candidates will know who the donors are) and also allows campaign committees to act explicitly on behalf of candidates in directly soliciting donations. If personal solicitations by candidates undermine "public confidence in judicial integrity," why do not these other practices create the same harm? The Court acknowledged that Florida does allow activities that might create some suspicion over whether judges are beholden to or favor donors, but concluded that "narrowly tailored" does not mean "perfectly tailored," and that the "First Amendment does not put a State to [an] all-or-nothing choice." For the Williams-Yulee majority, it was sufficient that Florida has targeted the "conduct most likely to undermine public confidence[,]" and that personal solicitations are "categorically different" from solicitations by campaign committees. The Court did not go to great lengths to explain this "categorical" difference, other than to say that while committee and personal solicitations may be "similar . . . in substance, a State may conclude that they present markedly different appearances to the public."

Importantly, though, the Court did not cite to, or seem to insist upon, any proof by the State that these two types of solicitations were viewed differently by the public. Indeed, when the Court said that a State "may conclude," it was using language most often associated with deferential review-where benefits-of-the-doubt about the real-world state of affairs are given to the government-not the language of truly strict scrutiny, in which the government must establish not just that its views are plausible, but that its views are grounded in actual fact.

3. Stare Decisis Is Often Not Very Powerful at the Court

The seemingly generous implementation of strict scrutiny brings up another important facet of Williams-Yulee-its tension with the most relevant Supreme Court case in the realm of judicial election regulation. There is, as one of the Williams-Yulee opinions put it, "only [one] prior case concerning speech restrictions on a candidate for judicial office"-the 2002 case of Republican Party of Minnesota v. White. And in that case the Court (in striking down Minnesota's judicial election regulation) applied a stricter version of strict scrutiny.

In White five Justices used the First Amendment to strike down a Minnesota law that prohibited candidates for judicial office from speaking out on controversial issues of the day. The law at issue prohibited a candidate for elected judicial office from "announc[ing] his or her views on disputed legal or political issues." The prohibition went beyond candidate "promises" and forbade, for example, a candidate from criticizing a past court decision and indicating a willingness to consider a different result in similar cases down the road.

Minnesota argued that it needed to regulate candidate speech to ensure that the public believes that judges are sufficiently open-minded about important matters that might come before them, an interest very similar to Florida's goal of "preserving public confidence in judicial integrity." But Justice Scalia's opinion for the majority in White rejected this justification for Minnesota's law because the scheme was woefully underinclusive, insofar as judicial candidates were not prohibited from voicing their views prior to the time they became declared candidates. The Court rejected the argument, made by dissenting Justices, that "statements made in an election campaign pose a special threat to open-mindedness because the candidate, when elected judge, will have a particular reluctance to contradict them." The Court said that the idea that judges feel particularly constrained by statements they make qua candidates is "not self-evidently true[,]" and thus cannot carry the day given the "burden [on the government] imposed by our strict scrutiny test to establish th[e] proposition that campaign statements are uniquely destructive of open-mindedness [or the appearance of open-mindedness]."

The tension between White and Williams-Yulee is clear. In the former, the State lost because it did not prove that campaign statements were "uniquely" destructive of the appearance of open-mindedness, but in the latter the State prevailed because it was allowed to "conclude" (without any proof) that personal solicitations "present markedly different" appearances to the public as compared to committee solicitations. Why Minnesota had to prove "unique" destruction of confidence whereas Florida could simply reasonably surmise "markedly different" problems of public perception is left unexplained.

Let me be clear here that I think the overall approach of Williams-Yulee is largely correct and that the analysis of the White majority was largely misguided. As I have written in law review articles and elsewhere, while the First Amendment protects one's right to speak about the bench, there is no right to to sit on it, and the Tenth Amendment gives states broad powers to regulate the process by which people become judges. The key point is not merely that judges are not supposed to be politicians; it is that throughout American history, we have often selected judges (but not legislators or chief executive officers) without the use of contested elections. And in these non-election processes, what would-be judges have said and done is held against them by government decisionmakers. Just as the president and the Senate certainly, and permissibly, may refuse to make someone a federal judge because of what that person has said, even though such refusals are undeniably "content-based" and indeed "viewpoint-based," and thus might, in other contexts, run afoul of basic First Amendment principles, a state should be generally available to deny judicial office to candidates who speak in ways that contradict certain judicial decorum norms set by the state. (There is the separate question, implicated in both White and Williams-Yulee, of whether the sanction for violating campaign rules can extend beyond mere disqualification for judicial office, which is a topic I save for another day.)

But my point here is not that Williams-Yulee's result is wrong-only that its application of strict scrutiny is not very authentic and that its leniency contradicts the approach in White.

4. Chief Justice Roberts Is no Clone of Chief Justice Rehnquist

How do we explain the tension between White and Williams-Yulee? The answer seems to rest largely on changes to the Court's personnel. White was a 5-4 case, with the majority consisting of Chief Justice Rehnquist and Justices O'Connor, Scalia, Kennedy, and Thomas. The dissenters were Justices Stevens, Souter, Ginsburg, and Breyer.

In Williams-Yulee, the remaining White dissenters (Ginsburg and Breyer) are (predictably) in the majority, and the remaining members of the White majority (Justices Scalia, Kennedy, and Thomas) are (predictably) in the dissent. Between White and Williams-Yulee, Justice Alito replaced Justice O'Connor, and voted the same way as we would have expected her to vote, and Justices Kagan and Sotomayor replaced Justices Stevens and Souter, and voted the same way as we would have expected them to vote. So far, so good-an even swap.

But Chief Justice Roberts, who replaced Chief Justice Rehnquist, did not follow in the footsteps of his predecessor here. So what was a 5-4 majority in favor of the First Amendment claimant in White became in Williams-Yulee a 5-4 majority in favor of the State. Chief Justice Roberts apparently has a different view of judicial elections (and the extent to which First Amendment protections for election-related speech apply to them) than his mentor and former boss. Whether there is a broader divergence between Chief Justice Roberts and his predecessor in First Amendment cases is a question that might be worthy of more attention now that the Roberts Court is finishing its first decade.

Cross-posted from Justia’s Verdict. Co-authored with Professor Alan Brownstein.

As we noted in a column for this site two weeks ago, state religious freedom restoration acts, or RFRAs, such as the recently amended Indiana religious liberty statute, have been criticized on the ground that they are intended to permit discrimination against gays, lesbians, and same-sex couples in the provision of goods and services. Given the intensity of this national controversy, we think it would be useful to take a step back-indeed, to take several steps back-and look at the historical background and evolution of the RFRA device. In this column, we focus not on any particular state statute but rather on three general topics: (1) the purpose of the earliest state RFRA laws and how that purpose relates to the goals of the more recently enacted and proposed legislation; (2) the virtues (and drawbacks) of enacting a general religious liberty statute as opposed to adopting religion specific accommodations on a case-by-case basis; and (3) the best way, in light of the current controversy about the conflict between state RFRA laws and anti-discrimination principles, to move forward when state legislatures consider these laws.

The Purpose of Early State RFRAs and What It Tells Us About the Recent Legislative Efforts

As we discussed in Part One, the Supreme Court, in 1990, decided the case of Employment Division v. Smith, a dispute involving the right of Native Americans to use the proscribed substance of peyote in their religious rituals. The Court ruled that neutral laws of general applicability are not subject to any rigorous scrutiny even when these laws have the effect of burdening religious practices. Unless the state targets religion-think of a law prohibiting Catholics from attending Mass-the Free Exercise Clause of the First Amendment simply does not require the state to explain or justify a law that has the effect of prohibiting religiously mandated practices or requiring the performance of religiously prohibited conduct.

The Smith decision came as a surprise to many, perhaps most, constitutional scholars. Based on prior cases, the parties to Smith had assumed that the Free Exercise Clause required, even in the context of neutral laws of general applicability, the government had to justify burdens on religious practice by showing that laws creating such burdens were narrowly tailored to accomplish compelling governmental interests. It is true that the Supreme Court, in applying this "strict scrutiny" narrow tailoring/compelling interest test had very rarely actually ruled in favor of a plaintiff asserting a free exercise claim against a general law. But it had often reached its conclusion by nuanced application of strict scrutiny, rather than rejection of the need for meaningful governmental justification altogether. Prior to Smith, lower courts could not summarily dismiss free exercise claims. After Smith, the door to the federal courts was, in effect, locked tight against free exercise claimants.

The Smith decision drew fire both from legal scholars and advocacy groups. In 1993, Congress enacted the federal Religious Freedom Restoration Act (RFRA), essentially to reinstate, as a matter of federal statute, the strict scrutiny religious liberty rights regime that individuals and institutions had previously understood to emanate from the Constitution itself. But in 1997, in City of Boerne v. Flores, the Supreme Court held that Congress exceed its enumerated powers in enacting RFRA insofar as RFRA applied to and regulated state and local governments. After Boerne, RFRA could be constitutionally applied only to burdens on religion created by the federal government.

This was the legal and political background against which several states considered the enactment of the first wave of state RFRA laws. It is important to recognize three conditions that characterize the consideration of state RFRA laws during this initial period in the late 1990s. First, support for or opposition to these laws did not correlate tightly to party affiliation. There was no doubt concern by some liberals about the application of state RFRAs to civil rights laws, but this concern was only part of the debate and did not cause legislators to be divided along party lines in their ultimate views on state RFRAs. In California, for example, in 1998, a state RFRA law passed both houses of the Democratic legislature, only to be vetoed by Republican Governor Pete Wilson.

Second, general concerns about the correctness of the Smith holding fueled the movement toward state RFRAs. Religious liberty proponents continued to believe and argue that free exercise rights should count for something if they were substantially burdened even by a neutral law of general applicability.

Third, the arguments in favor of state RFRAs were not grounded just in abstractions; they were nested in actual cases and real-world narratives. A pair of real-life settings received particular attention. One was land-use regulation. Religious congregations, it was argued, often found it extremely difficult to develop land to construct new houses of worship because of restrictive state and local zoning laws. Many towns didn't seem to want new venues of worship in residential areas, or commercial districts, or even in agricultural zones. And minority faiths seemed to bear the brunt of these regulatory restrictions. The other narrative involved the religious freedom of prison inmates. It was widely believed that state prison authorities imposed relatively arbitrary burdens on the ability of inmates to engage in worship or other religious activities.

The first and third of these conditions no longer exist today. As to the first, religious liberty legislation is far too often a partisan political issue at this moment, with Republicans favoring state RFRAs and Democrats opposing them.

And, importantly, as to the third, the pair of persuasive narratives for adopting a state RFRA-the burdens created by state and local land use regulations on congregations trying to develop land for a new house of worship and the difficulties state prison inmates experienced in engaging in religious worship and exercise-were effectively dealt with by federal legislation. In 2000, Congress enacted the Religious Land Use and Institutionalized Persons Act (RLUIPA). RLUIPA requires state and local governments to justify under rigorous review land use regulations or decisions that substantially burden the use of land for religious purposes and prison regulations and decisions that substantially interfere with the ability of inmates to engage in religious worship or otherwise follow the dictates of their faith. Because RLUIPA invokes Congress's Spending Clause power to attach conditions to federal funding-and because virtually all state and local governments and prisons depend on federal funding-RLUIPA has been upheld and applied by countless lower federal courts and (in the context of the prison provisions) the Supreme Court.

Recent state RFRA laws and proposals can still be justified by the second backdrop condition animating the first generation of state RFRAs-the abstract idea that Smith created a gap in the protection of religious liberty, and that religious activity deserves to be protected to some extent against even neutral laws of general applicability. But because, other than the land-use and prison settings, there are no easily described categories of state regulatory activity that burden religion in ways most people find problematic, a modern state RFRA might seem like a solution in search of a problem. Indeed, the only unifying narrative that describes a general problem, as opposed to isolated cases, to which modern RFRAs might be directed is the narrative grounded in religious objections to same-sex marriage and the claims for exemptions from civil rights regulations that prohibit discrimination on the basis of sexual orientation.

This is the crux of the problem. Legislators and governors who argue that they support a state RFRA law today for reasons that have nothing to do with discrimination related to same-sex marriage have a difficult time persuading anyone of their position because there are no religious liberty narratives involving significant real-world areas of concern other than civil rights laws. The original, principled basis for enacting state RFRA laws still exists, and state RFRAs certainly may be of value to religious individuals or institutions in occasional varied circumstances-religious burdens do arise outside of land-use and prison contexts as we demonstrated with some hypothetical examples at the end of Part One of this series-but there is no well-organized storyline here that can be easily understood and valued. In other words, because, after RLUIPA, the contexts in which state RFRAs might provide needed protection do not fall into any easy-to-define or easy-to-predict categories of regulation, avoiding antidiscrimination laws is the primary narrative that is still left standing. It is the one that most people see. And, to be frank, it certainly appears to be the primary motivation for the introduction of new RFRA bills in state legislatures these days.

A General Religious Liberty Statute Versus Religion-Specific Accommodations on a Case-by-Case Basis

Assuming that some exemptions for religious activity outside of the land-use and prison arenas may be worthwhile, the question becomes whether a state statute (e.g., a RFRA) is the best way to facilitate them. Another way of framing the issue is, given that some religious exemptions will be recognized by government, whether we are better off determining when exemptions should be granted by having the more political branches of government evaluate practice- or sect-specific requests for accommodation, or whether it would be preferable to enact a general religious liberty statute, like a state RFRA, and shift the task of determining when an exemption is appropriate to the judiciary. We think general religious liberty statutes have some important virtues over religious practice- or sect-specific accommodations.

First, the general religious liberty statute is, by definition, general. It seeks formally to apply the same standard to all faiths. Thus, a religious person's ability to obtain an exemption will not, in theory at least, depend on his or her ability to influence the political branches of government. It is true that judges, like legislators, may also be unfamiliar with or unsympathetic to religious minorities. Still, under a general religious liberty statute, a minority faith with insufficient muscle to achieve an accommodation through political channels has an additional forum where its claims can be heard-a court of law.

Second, the business of obtaining sect- or practice-specific accommodations has other serious drawbacks. Restricting religious exemptions to the political branches of government politicizes religion. The freedom to practice one's faith becomes a benefit controlled by the government. Accordingly, religious groups have to organize politically as religious groups to obtain the exemptions their faith requires.

Third, and related, a system in which all accommodations are political actions requires religious individuals and groups to spend their political capital on freedoms that should be theirs as of right. This system operates like a political tax on religion.

Fourth, if the ability to practice one's faith depends on a religious group's political power in a jurisdiction, we create an incentive for religious people to live in communities where there are a sufficiently large number of co-religionists to influence the government. A legal regime that promotes the segregation of communities along religious lines is problematic and much less desirable than a regime that facilitates the religious integration of our communities.

Of course, there are problems with general religious liberty statutes as well. The standard of review to be applied by courts in these laws is intrinsically subjective, value-laden and unpredictable. No one can really be sure how a given court will interpret and apply the law to the facts of any given case. Accordingly, the protection provided to religious liberty may turn out to be much narrower or much broader than the community anticipated when it enacted the law. In theory these statutes can be amended to cure wrongly decided cases, but there is no guarantee that the political branches of government will be capable of effectively monitoring and responding to errant RFRA decisions by courts.

Moreover, the indeterminacy inherent in these laws means that, at least initially and in all cases of first impression, they will provide little guidance either to potential defendants or to plaintiffs. In the context of anti-discrimination laws, uncertainty imposes serious burdens on all the relevant parties. Service providers do not know if they are permitted to deny services for same-sex weddings, for example, because of their religious objections to such ceremonies. And same-sex couples lack the security of knowing that they cannot be denied the services they seek when they attempt to patronize a provider of wedding services.

While we recognize that reasonable people can disagree on this point, we think on balance there are legitimate reasons for a state to consider enacting a state RFRA law. But that does not mean that we think the RFRA law should operate to provide exemptions in all cases in which religious exercise is substantially burdened by law.

What is the Best Way for State Legislatures to Balance State RFRAs and Anti-Discrimination Principles?

Church-state scholars generally agree that most RFRA challenges to civil rights laws governing for-profit economic activity will and should be unsuccessful. The state has a compelling state interest in protecting members of particular classes against discrimination in the workplace and in places of public accommodation. And conventional civil rights laws are the least restrictive means available to accomplish this egalitarian goal. Still, no one is certain that all RFRA claims against regulations prohibiting discrimination will fail. Nor is there agreement as to which claims, if any, deserve to succeed.

Because RFRA laws are unlikely to provide any kind of expansive protection to discrimination in employment or public accommodations based on religious beliefs, an obvious solution to the controversy surrounding these laws would be to enact a civil rights carve-out that limits the scope of the RFRA legislation. Indiana amended its RFRA law to provide explicitly that the law does not authorize, or establish a defense for, discrimination in employment or places of public accommodation. Such a civil rights carve-out would make the RFRA law available to protect religious liberty in in various idiosyncratic circumstances in which general laws unnecessarily burden religious practice, but would preclude any possibility that the law would undermine the enforcement of anti-discrimination regulations.

The argument against a civil rights carve-out is that it could carve out too many RFRA claims. Many proponents of state RFRAs argue that there are at least a limited number of situations in which religious exemptions to some civil rights laws are justified, and yet these claims would be excluded from protection under a general civil rights carve-out. These arguments often focus on caterers, bakers, florists and photographers who provide goods and services for wedding ceremonies and receptions, but the arguments are not limited to these commercial activities.

We think the appropriate response to these concerns is straightforward. In addition to adopting a broad civil rights carve-out from the state RFRA, the state could negotiate explicit exemptions-exceptions to the carve-out, if you will-to cover the limited number of situations in which faith-based discrimination might deserve to be protected against civil rights laws. From a policy perspective, this approach would have several advantages. It would provide more clarity than a generic state RFRA. It would guarantee religious exemptions to civil rights laws in specific circumstances where they were thought to be particularly justified. It would avoid any concern that the law would be interpreted too broadly to protect discrimination in inappropriate circumstances. And it would allow a state RFRA to be adopted to protect religious liberty in all of the situations that do not involve discrimination in violation of civil rights laws.

Our suggested course of action may be challenged, however, by the argument that such negotiations in the legislature about the particular exceptions to a civil rights carve-out would be futile. The two sides debating religious liberty and gay rights issues are so polarized that they would never agree on explicit limited exemptions. We are unconvinced that this will always be the case-particularly if states that currently do not protect gays and lesbians or same-sex couples from discrimination bring legislation prohibiting discrimination based on sexual orientation and identity to the bargaining table. Working out what the specific exemptions for religion-based discrimination will undoubtedly be hard political work. But that is no reason not to engage in the attempt.

Cross-posted from Justia's Verdict. Part one in a two-part series of columns. Co-authored with Alan Brownstein.

Over the past month or two, religious accommodation laws that have been enacted or proposed by states have attracted much attention in the media and among legal analysts. Such state laws are often called Religious Freedom Restoration Acts, or RFRAs-named and patterned after the federal RRFA adopted by Congress after the Supreme Court's 1990 decision in Employment Division v. Smith, where the Court interpreted the First Amendment free exercise protection narrowly to reject a claim by Native Americans to use the prohibited drug peyote for religious purposes. RFRAs require that before government is allowed to impose a substantial burden on the practice of someone's religion, the government must have a compelling objective that cannot be accomplished by any narrower means for doing so. State RFRAs have been around in some states for a few decades, but this spring saw a new round of state legislative activity in places like Indiana and Arkansas, presumably triggered by the anticipated tension between the tenets of some religions and the ruling most analysts expect the U.S. Supreme Court to render this summer making clear that the legal institution of marriage cannot be denied to same-sex couples.

Other Verdict columnists have already offered insights and arguments about the best way to understand and interpret state RFRAs. In this two-part series, we offer our own take on the state RFRA movement and how best to incorporate it into a nation dedicated to free religious exercise and separation of church and state at once. In Part One, in the space below, we offer some reactions to the doctrinal analyses presented in a recent essay by Verdict columnist Michael Dorf. In Part Two, in a few weeks, we widen the focus to examine more fundamentally how and when state RFRAs came about and what their origin should mean for how they should be implemented.

Mike Dorf's Analysis of State RFRAs in the Context of Private Litigation

Mike Dorf's elegant doctrinal analysis of state RFRAs focuses on whether these laws "should apply in private litigation [i.e., litigation in which neither party is a government entity] if the statute is silent on the matter." Mike offers a couple of arguments for why state RFRAs perhaps ought not to apply to private lawsuits altogether. His first argument begins with a reminder that RFRAs are designed to "restore" the "constitutional right to free exercise of religion that was weakened by the U.S. Supreme Court in its 1990 peyote decision. Because a RFRA restores a constitutional right that applies only against the government, it is natural to assume that a RFRA should be available only in litigation against the government."

But, as Mike rightly points out, oftentimes constitutional rights are at stake and vindicated in cases in which the government is not a party, but in which a party is using some law or policy the government has adopted as the basis for its legal position. So, for example, when a public-figure plaintiff sues a magazine under the state tort law of defamation, the defendant can properly invoke the First Amendment as a defense, even though the plaintiff is a private individual rather than the government, because the plaintiff is relying on state-adopted tort law for his claim. It is the state, through the creation of its tort law, that is effectively burdening the defendant's speech.

Or, as in another example Mike offers, if a state passes an alimony law that treats men and women unequally, such a law can be challenged in a lawsuit between a divorcing husband and wife, even though the state is not a party, because one of the parties is so directly invoking the state law as the basis for asking a court to do something.

Mike properly acknowledges that even in the context of religion, a state's fingerprints can be all over a burden imposed on someone's religion, even if the state is not doing the litigating. So, for instance, if a state gives a landowner's neighbor a right to veto the landowner's decision to expand his building, and a church that wants to expand is blocked by a vetoing neighbor, the church might seek to invoke the free exercise of religion as a basis for resisting the veto, even if the opposing party in the lawsuit is the neighbor to whom the state has given the veto right instead of the state agency itself.

Does Private Litigation Under a RFRA Implicate State Action in a Way Different From Cases in Which Government Is a Party?

After all this, however, Mike argues that the state's involvement in RFRA cases is distinct in a way that perhaps argues against allowing state RFRAs to be invoked in private litigation. Says Mike, about the examples he offered earlier: "When [a defamation defendant] invoked the freedom of the press against [the defamation plaintiff], it objected that the [state] tort rule was defective in permitting a public figure to prevail [under a standard] that afforded insufficient protection for free speech. . . [And] [w]hen [a husband] resisted his alimony obligation, he complained that the [state] statute favoring women over men denied him equal protection of the laws. In these, and many other situations, the party invoking a rights provision in private litigation argues that some legal rule or standard violates his, her or its own rights. In contrast, a RFRA claim does not challenge any rule or standard."

Here is where we think we disagree with Mike. A RFRA claim does challenge a rule or standard-the rule or standard on which the private party opposing the religious claimant is relying in the private litigation. The fact that the right a RFRA claimant seeks to invoke is a statutory (RFRA-created) right to religious accommodation, rather than a constitutional right (such as the right to free speech or equal protection), is beside the point; remember, RFRAs are designed to "restore," by statute, the liberties previously recognized under the First Amendment's Free Exercise Clause. The RFRA claimant has been conferred a right, just as much as a free speech or equal protection claimant has been. And state law, it is alleged in RFRA cases, is protecting the other party's ability to violate that right-by substantially burdening the religious claimant's exercise of his or her religion.

Mike's instinct that a RFRA claimant is not alleging that any state law creating a burden is "defective" is understandable but, we think, wrong. A law challenged by a RFRA claimant is indeed "defective" in the legally technical but important sense that it (allegedly) fails to adequately accommodate religion, which is what the RFRA seeks to guarantee. In the defamation case alluded to above, state tort law wasn't defective in any a priori sense; it was defective only in the sense that it failed to sufficiently accommodate free speech. And RFRA claimants make the same claim as to religion.

Indeed, the example Mike offers concerning the neighbor's veto over land-use decisions seems to illustrate our point. If a church's plans to expand are blocked by a zoning board, clearly the church could invoke both the First Amendment prior to 1990, and a state RFRA nowadays. The same should be true if the opposing party is not the zoning board, but the vetoing neighbor. The law giving the neighbor veto power is defective not in a generic sense, but only in the sense that it may have the effect of frustrating religious freedom. Yet it ought not to matter whether the opposing party is the government or the neighbor himself, or whether the claim is brought under the First Amendment (before it had been watered down) or a RFRA (that seeks to reclaim the undiluted religious right).

We think our analysis makes sense in part because a state can (and often does) elect to have a lot of different kinds of laws enforced through private causes of action-and when it chooses to do so we often find there to be "state action" in the enforcement. The Supreme Court's willingness to find state action involves several factors and seems to vary depending on the particular freedoms that are at issue. We note that the Court has taken a particularly expansive approach to state action in interpreting the Establishment Clause, and it would not be unreasonable to argue that a similarly expansive understanding of state action should apply Free Exercise values. And if there is state action, if the burden would be sufficient to trigger free exercise review if the state itself enforced the law, why should it make any difference if the law is enforced by a private party?

What About Third-Party Burdens?

Mike's second argument for perhaps not applying state RFRAs to private litigation arises from the fact that in all private litigation, accommodating religion creates "the potential for substantially burdening a third party." And the Supreme Court, in the recent Hobby Lobby decision and elsewhere, has given indications that accommodating religion when such accommodation takes the form of inconveniencing government is one thing, but religious accommodations that impose on third parties may be another thing entirely.

Like Mike, we think third-party burdens ought to figure prominently in any application of state RFRAs. But we are not sure a prophylactic rule prohibiting invocation of a RFRA in all private litigation is necessary to properly take account of third parties. Because state law may allow private individuals who don't suffer much, if any, injury to be in litigation against religious adherents (remember that state courts are not limited by the Constitution's Article III standing rules), and because some third-party injuries may be of such a nature that avoiding them cannot reasonably be thought to be a compelling government interest, we think the better course is not to categorically reject RFRA claims in private litigation, but to examine any third-party burdens on a case-by-case basis. When racial or gender discrimination is at issue, the third-party costs will justify denying the accommodation. But imagine the following two hypotheticals:

1. Suppose a municipal stadium district has a rule that says no one can wear hats taller than 5 inches to sporting events, because people's views get blocked, and allows for a private right of action in small claims court by aggrieved persons. Suppose someone wears a turban to a football game, and gets sued for $500 by another fan seated behind him who had to stand up more often to see the action.
2. Or suppose a City bans discrimination in the provision of goods and services against people who openly display tattoos. A religious small businessperson who runs his business out of his home declines to serve a patron because the patron refuses to cover up a sexist tattoo on his upper arm, and display of such an image in the home violates the religious tenets of the businessperson. The aggrieved customer sues.

In both of these examples, accommodating religion does create some state-recognized burdens on third parties. But are they the kinds of burdens that would justify a flat, prophylactic rule prohibiting invocation of a RFRA in all private litigation? We are not yet sold on that. Thus, if a state RFRA does not by its terms prevent its application to private litigation (and, of course, every RFRA must be interpreted in light of its own language, read in the context of the entire statute), we think the better course may be to examine each such private litigation case on an individual basis, to look carefully at the extent of state action and third-party burdens.

In Part Two of this series, we locate state RFRAs in a larger historical and doctrinal context, and offer some thoughts on how to give meaning to state RFRAs while avoiding some of the externalities and complications with which Mike is properly concerned.

Cross-posted from Justia's Verdict.

In today's column, I analyze the Supreme Court oral argument held a few weeks ago in Walker v. Texas Division, Sons of Confederate Veterans, a case involving the First Amendment and Texas's regulation of license plates. Motor vehicles registered in the State of Texas must display a state-sanctioned license plate. Most vehicles use a standard-issue Texas plate that has a simple no-frills design and displays a random series of letters and numbers. Texas, like many other states, also permits individuals to submit personalized, or vanity, plates in which the numbers and letters on the plate form a message (such as "HOTSTUFF," a hypothetical example Justice Scalia used at oral argument).

In addition, Texas permits what are called "specialty" license plates, in which the overall design of the plate (but not the sequence of numbers and letters), is custom-made and might contain symbols, colors and other visual matter that is more elaborate than the relatively plain design of the standard-issue plates. Specialty designs may be adopted by the Texas legislature or proposed by private individuals or organizations. Specialty plate designs that come from outside the legislature must be approved (as must personalized vanity plates) by the Texas Department of Motor Vehicles Board, and the Board by law "may refuse to create a new specialty license plate if the design might be offensive to any member of the public" (a term that Texas authorities construe as meaning offensive to a significant segment of the public.) At least some specialized designs, once approved, can be used by members of the general public. As of a month ago, there were about 450 specialty designs that had been approved in Texas, around 250 of which are usable by the public. Although the majority of license plates in Texas are the plain-vanilla non-specialty plates, it is not uncommon on the Texas roadways to see license plates that make use of one of the approved specialty designs.

Applicants who seek approval of specialized plate designs must pay thousands of dollars to have their designs considered, and people who use the already approved designs pay for the privilege, the proceeds going to various state agencies.

The Texas Sons of Confederate Veterans (SCV) is a nonprofit organization dedicated to preserving the memory and reputation of Confederate soldiers. SCV applied for a specialty license plate that featured the SCV logo, "which is a Confederate battle flag framed on all four sides by the words 'Sons of Confederate Veterans 1896.'"

When this design was rejected by the Motor Vehicles Board (one of only a dozen or so designs that have been rejected), SCV sued, arguing (successfully in the lower court) that the State's decision to reject the design on the ground that the content of the design-in particular, the depiction of the Confederate flag-might be offensive to some observers constituted impermissible content- or viewpoint-based regulation of expressive activity insofar as the specialty license plate, while State property, is akin to a forum for speech that the government has created and opened up to people to use to express themselves. Texas, for its part, argues that because the State owns all license plates, and because the State of Texas name appears on all plates, including specialty plates, any expression on license plates constitutes "government speech" or at the very least a hybrid of government speech and private speech. Because the government is a (if not the) speaker in this setting, Texas argues, it necessarily has the authority to accept and reject whichever messages it chooses.

The case raises many fascinating and complex constitutional issues-far too many to meaningfully address in a single column. But in the space below, I use three particular kinds of questions that Justices asked at oral argument to illuminate important and often misunderstood aspects of First Amendment doctrine.

Less Can Be More (Important) Under the First Amendment

Let us first consider Justice Kennedy's questioning of the SCV lawyer. One of the things Justice Kennedy pointed out is that if Texas is not permitted to exclude Confederate flags (or Swastikas, or other potentially objectionable material) from license plates, it will almost certainly choose to abandon the specialty (and also the vanity) license plate design program altogether, and simply use old-fashioned, plain vanilla license plates. The result, said Justice Kennedy, is that we would end up with less, rather than more, speech, because individual expression that is currently taking place on specialty or vanity plates would no longer be permitted, and people would be forced to resort to things like bumper stickers, which they may not like or make use of as much as specialty plates. "If you prevail," Justice Kennedy asked SCV's lawyer, "you are going to prevent a lot of Texans from conveying a message. . . . So in a way, your argument curtails speech?"

Justice Kennedy's question is actually a profound but rarely explored one, in that the First Amendment's aversion to content- and viewpoint-based laws may indeed sometimes lead government to enact content-neutral counterpart laws that, quantitatively speaking, restrict far more speech. For example, a law that says "no pro-life rallies in the park after 6pm" is clearly unconstitutional, because it regulates speech on a matter of public concern in a traditional public forum in a viewpoint-based way. But if such a law is replaced with a law (that very well might be upheld) that simply prohibits all rallies in the park after 6pm-a so-called content-neutral regulation of time, place or manner-the result could be an even greater overall reduction in speech.

Of course, it is possible that by forcing government to regulate in a content-neutral way, we may actually make it harder for government to regulate speech at all, so that the end result could actually be an increase in the aggregate level of speech. In the example I gave above, perhaps it would politically difficult to pass a law prohibiting all rallies in the park after 6pm (because many kinds of groups may want to hold rallies, and overcoming the political opposition of all of these groups-as opposed to the merely the pro-life advocates-may not be feasible). If that is true, then striking down the law prohibiting pro-life rallies after 6pm will, in fact, increase rather than reduce the amount of speech.

But oftentimes (as in the SCV case) striking down a law on First Amendment grounds may in fact lead to less speech, but it still can be the right constitutional thing to do. The fact that sometimes we invalidate laws in ways that will create less speech overall tells us that maximizing the aggregate quantum of private speech is not the only thing the First Amendment is concerned with. Preventing the government from distorting the debate, by disabling some points of view, or by locking in majoritarian preferences (as is often the case when "offensive" speech is disfavored) is also an important objective. So too is making individuals feel that government respects them and does not act paternalistically and treat them the way parents treat children by telling them what topics they should be focusing on.

What's Good for the Goose. . . .

A second line of questioning of SCV's lawyer, this time by Justice Sotomayor, concerned whether the State should be given the same kind of free speech respect as individuals enjoy. Justice Sotomayor pointed out that that in the Court's most famous license-plate case to date, Wooley v. Maynard, the majority struck down a requirement that New Hampshire drivers make use of a state-issued license plate bearing the State's message "Live Free or Die." Justice Sotomayor then asked: "In Wooley we said we can't compel the individual[s] to put something on their plates that they disagree with . . . Why isn't the reverse true for the government [if it doesn't want to be associated with the Confederate flag]?"

Justice Sotomayor's symmetry instinct (which assumed arguendo that the Texas specialty license plate regime represents at least a hybrid of government and private speech) is very interesting but ultimately unpersuasive, to me at least. There are lots of constitutional rules that protect individuals that do not protect government in a symmetrical way. For example, a criminal defendant is entitled to have access to all exculpatory evidence in the government's possession, but the government is not entitled to all incriminating evidence in the defendant's possession, even though both sides are trying equally hard to prove their case.

I think there is asymmetry here as well. Even though the government can operate as a speaker, it is not a specific beneficiary of the First Amendment, and certainly shouldn't enjoy all the same First Amendment protections individuals (like the individuals who litigated in Wooley) do. Ultimately, the reasons the drivers in Wooley could not be forced to bear the State's message were rooted in individual dignity and autonomy aspects of the First Amendment. Institutional and organizational actors, as opposed to individuals, can be forced to be a vehicle for government messages and are relegated to engaging in counter-speech as a way of distancing themselves from any government message they don't like. This was true in Rumsfeld v. Forum for Academic and Institutional Rights (FAIR), a case last decade that upheld (9-0) a federal law that required law schools to allow military recruiters onto campus facilities to recruit students, notwithstanding the law schools' opposition to the then-existing policy of the military to discriminate against gays and lesbians ("Don't Ask, Don't Tell"). Like law schools, the State of Texas does not have the same kind of dignity and autonomy attributes that individual motorists have, and so (even granting that Texas has the authority to act as a speaker) requiring Texas to live with the private message on specialty plates and disclaim any endorsement of the message or design on a specialty plate by adding something like "Views on this license plate do not reflect the views of the State" does not violate the Constitution the same way requiring individuals to promulgate such disclaimers would.

The Relevance (or Not?) of a Profit Motive

A third interesting exchange involved the overridingly important question of whether the specialty plates can properly be thought of as pure (or at least hybrid) government speech at all. The State's lawyer argued that the fact that the government has retained the right to veto all specialty designs from the get-go makes this a government speech case, but that factor standing alone surely cannot be dispositive. If a public airport withheld for itself the power to ban any leaflets whose message it found unattractive, that would not justify its excluding leaflets in favor of affirmative action while permitting leaflets against it. Control is, as many Justices pointed out, a circular kind of argument about government power. Deciding what is and is not government speech is much more complicated than that.

One potential factor was mentioned by Chief Justice Roberts a few times, and that is the profit motive by the State. Why, he asked, should we view these specialty plates as government expression at all when government's real goal here was not to raise awareness (about anything) but to raise money? This, too, is an interesting instinct. As with Justice Sotomayor's question, if we analogize to private individual speech, the government fares better; the fact that a private individual or corporation is motivated by a desire to make a profit does not make his/its expression any less constitutionally valuable: the New York Times newspaper represents classic First Amendment speech even though it is published in order to make money.

But as was true for Justice Sotomayor's symmetry argument, here too I am not sure we should treat the government the same as individuals. It does seem a bit untoward that the State would raise revenue by charging people thousands of dollars for the privilege of submitting license plate designs, and then reject those whose content it doesn't like. The idea that the State was (mis)using the specialty-design applicants, and the moneys they paid, for its own monetary gain was one of the most sympathetic aspects of the SCV's case, which was otherwise not very sympathetic given that the Confederate flag has historically been tightly associated with slavery and insurrection (not to mention the fact that SCV's lawyer took the position that the State could not, consistent with the First Amendment, reject designs with swastikas on them.)

The opinions that emerge from this case in the coming months could be very interesting.

Cross-posted from Justia's Verdict.

The U.S. Supreme Court last week heard oral arguments in an important case involving federalism and election regulation, Arizona Legislature v. Arizona Independent Redistricting Commission. As I have explained in a two-part series of columns (beginning with this one), Arizona is one of only two states (California being the other) where voters—responding to state legislatures’ tendency to engage in problematic gerrymandering—passed an initiative giving the job of drawing congressional districts to an independent redistricting commission (IRC) instead of the regular state legislature. The elected Arizona legislature brought a lawsuit and appealed the lower court ruling to the Court, arguing that the so-called Elections Clause, Article I, section 4 of the federal Constitution, which gives power to undertake districting in the first instance to the “legislature” of each state, prevents the people of a state from divesting the elected state legislature of district-drawing power.

One Generally (and Recurringly) Surprising Aspect: Federalism Inversion

Many aspects of the oral argument weren’t shocking. I continue to believe, for reasons I explained in the earlier column, that the challenge to the IRC is flawed, but many analysts anticipated that the more conservative Justices would be sympathetic to the arguments made by the elected legislature, which is being represented by former Bush Administration Solicitor General Paul Clement, and these Justices did seem to be. On the other hand, some liberal Justices generally seemed more receptive to the arguments in favor of the IRC, made by its lawyer, former Clinton Administration Solicitor General Seth Waxman. Though expected by Court watchers, this coalitional breakdown is itself surprising in at least one historical respect: the liberal Justices seem more inclined to favor “states’ rights” by giving states latitude to experiment with different modes of district drawing, while the conservative Justices seem disinclined to permit states free reign. In the 1980s and 1990s, conservative Justices were the ones who generally thought the federal Constitution allowed for broad state experimentation, and the liberals thought that states did not have as much running room. But beginning with Bush v. Gore, it is now hard to know how the conservative/liberal framework (which is, of course, overly simple but nonetheless somewhat useful) maps onto federalism matters. We see the same potential complexity in the pending Obamacare case, King v. Burwell, where liberal rather than conservative Justices will likely construe the Affordable Care Act (Obamacare) in a particular way so as to avoid imposing costs on states that might not have been fully aware of the potential consequences of their decisions not to set up their own healthcare exchanges.

A Surprising Inattention to Standing

In a few other respects, though, the oral argument in the Arizona case was surprising by any standard. For one thing, the Court included in the grant of review a question of whether the elected legislature had standing to sue (and the matter was briefed), yet the Justices asked no questions of Mr. Clement about whether his client did enjoy standing. There were some standing questions asked of the lawyer representing the United States (which supported the IRC as an amicus in the case) when he said he thought the elected legislature lacked standing, but one would think that if there is a question raised in a case about the plaintiff’s standing, the Justices would ask the lawyer representing the plaintiff to explain why the Court can hear the case. The burden to establish standing is on the party seeking to invoke federal judicial power, and yet the Justices gave Mr. Clement a pass on this in oral argument, even though the standing question is far from easy. (For an explanation of why the standing question is a complex one, readers may want to consult an earlier Verdict column focusing on that question).

And when they did engage in the standing analysis at all, the Justices seemed not to know what they themselves had said in past standing cases, including a particularly relevant one. Justice Kennedy, in questioning the lawyer for the federal government, intimated that the Court’s cases do not say that just because another plaintiff might be a better candidate to bring a lawsuit, the Court should deny standing to the party actually in front of it. And the federal government’s lawyer acknowledged that the Court often says close to the opposite—that the “even if [the Court’s rejection of standing in a given case] would mean no one would have standing to sue, that’s not a reason to find standing.” But what both of them seemed to forget is that in the standing case most germane to the Arizona legislature’s dispute, Raines v. Byrd—where the Court denied standing to members of Congress in a case challenging the Line Item Veto Act—the majority did suggest that whether another party might be a better candidate for standing was a factor that might cut against standing for the members of Congress. As Chief Justice Rehnquist’s opinion put it: “We also note that our conclusion . . . [that Congresspersons lack standing does not] foreclose[] the Act from constitutional challenge (by someone who suffers judicially cognizable injury as a result of the Act). Whether the case would be different if . . . [this] circumstance[] were different we need not now decide.” The clear implication of that last sentence is that the existence of a “better” candidate for standing may very well affect whether the Court is willing to stretch standing doctrine for the sake of the plaintiff actually before the Court.

A Surprisingly (and Problematically) Narrow View of Past Cases and U.S. History

This isn’t the only passage from previous cases that the Justices seemed to have forgotten they said. Justices Kennedy and Scalia repeatedly pressed Mr. Waxman for any instances elsewhere in the Constitution in which the Court had indicated that the word “legislature” was not necessarily a limited reference to the elected representative legislature, but could include the people themselves—the lawmaking authority. While Mr. Waxman had no clear answer, one thing he could have said is that Justice Scalia himself has intimated that “legislature” might in some constitutional contexts be read to include the people. In Salazar v. Colorado, in 2003, a case involving whether Colorado could involve state courts in discharging the power that Article II of the Constitution gives to “the legislature” of each state to prescribe the manner in which presidential electors shall be selected, Justice Scalia (along with Justice Thomas) joined an opinion, dissenting from a denial of certiorari, which said: “Conspicuously absent from the Colorado lawmaking regime, under the Supreme Court of Colorado’s construction of the Colorado Constitution to include state-court orders as part of the lawmaking, is participation in the process by a body representing the people, or the people themselves in a referendum” (emphasis added).

The Justices seemed somewhat forgetful not just of past statements in cases, but also of American history, in particular, the path by which U.S. senators came to be directly elected by the people. The text of the original Constitution (prior to the Seventeenth Amendment in 1913) provided that senators should be picked by the “legislature” of each state. Justice Kennedy, trying to draw a sharp distinction between the “legislature” and the people of the state based on this textual feature of the 1787 document, observed at argument that “until 1913, for close to a hundred years, many States wanted to have direct election of the senators . . . and not one State, not one State, displaced the legislature. It took the Seventeenth Amendment to do that. . . It seems to me that [for this reason] history works against the IRC.”

With all due respect, Justice Kennedy’s historical account here is extremely and problematically simplistic, in that many states did effectively, to use his term, “displace[] the legislature” in picking U.S. senators. Beginning in the mid-1800s, state-level political parties and organizations sought ways to involve the people more directly in selecting senators, and were devising increasingly effective ways to limit state legislators’ discretion in their choice of federal senators. What evolved into the most sophisticated approach, the so-called “Oregon Plan” (or Scheme), began simply as an opportunity for state legislative candidates to formally pledge to follow the will of the voters, as expressed through an advisory popular election, when it came time to pick the next federal senator. The pledges were considered merely moral at first. But as other states began to follow Oregon’s lead, more creative and more coercive devices were employed. Nebraska, for example, pioneered a “scarlet letter” approach, in which elected legislators who broke the pledge they took as state legislative candidates were burdened with a ballot notation to that effect in the event they sought state legislative reelection. Other states followed suit, crafting variations on the Oregon and Nebraska devices to suit their local needs. Oregon voters ultimately adopted a state constitutional amendment that, as a matter of state law, legally bound state legislators to select the U.S. Senate candidates who were most popular among state voters. By 1912, when the U.S. Senate approved the Seventeenth Amendment, nearly sixty percent of the senators were already selected by some means of direct election (and thus had nothing to fear from it). For this reason, it seems likely that even without ratification of the Seventeenth Amendment, direct election would in fact be with us today in most, if not all, states. In reality then, the Seventeenth Amendment was a formalizing final step in an evolutionary process.

Of course, the Oregon state constitutional provision binding state legislators, the Nebraska scarlet letter devices, and the somewhat similar measures from other states were never litigated in the U.S. Supreme Court or in lower courts. Yet that fact may itself be telling. Does Justice Kennedy think that these devices were in fact unconstitutional because they improperly deprived the elected legislatures of power Article I gave to them? If so, would Justice Kennedy be prepared to call into question the legitimacy of the senators elected from all the states that employed such devices for over a decade? And the actions taken by these Senates? And if not, doesn’t this historical episode support the Arizona electorate and its desire to experiment via the IRC?

In deciding what the word “legislature” in the Constitution means, in Article I and elsewhere—and whether that term can be read to include the people themselves—the Court should, at a minimum, be more careful and sophisticated in taking account of what the Court or various of its Justices have said, and what the full historical record of American democracy reveals.

Co-authored with Dean Kevin R. Johnson. Cross-posted from Justia's Verdict.

Over the next month or two, tens of thousands of admitted applicants will make decisions about which law schools to attend. One tool that many will no doubt use to guide their decisions is the annual U.S. News & World Report rankings, which will be released in a little over a week. Many analysts criticize the methodology (or various aspects of it) that U.S. News employs to rate law schools (and some folks doubt whether all the nation's law schools could ever be meaningfully graded according to any single set of criteria.) But, for the time being at least, U.S. News remains the most looked-at, and seemingly influential, ranking system out there. For that reason, in the space below we offer-based on our collective experience in both evaluating other law schools and having our own law school evaluated-five pieces of advice for making the most sophisticated use of the rankings that U.S. News is poised to unveil.

#1: The Importance of Trends: Remember That Each Year's Rankings Capture a Snapshot in Time

The rankings that are set for release on March 10 present a great deal of raw and processed information, but they data they contain-and the bottom-line rankings they assign-represent only a snapshot in time. Any sensible consumer of the rankings should look not just at one year's result, but at a longer track record, perhaps attaching more weight to a five-year average rather than to any single year's numbers.

To be sure, sometimes there is, as to a particular law school or type of law school, a clear trend line-in particular components within the ranking or as to the bottom-line performance - and it may be important to try to discern what accounts for any such consistent assent or decline. More commonly, a school may bounce around somewhat because of short-term factors, such as a bad year in passing the bar and/or placing graduates in jobs, or an anomalous drop in application volume or quality due to some administrative gaffe or regional downturn. Such volatility is itself a basis on which the U.S. News rankings are often criticized-how much could a school's overall quality really change within the space of a year?-but taking a somewhat longer view may partially address that criticism and make the bottom-line ratings more meaningful.

In looking at changes over time, it is important to realize that certain parts of the U.S. News evaluations very rarely move much from year to year. This would include a school's reputation rank among other law professors who are surveyed (which accounts for 25% of a school's overall ranking) and its reputation rank among lawyers and judges who are polled (which accounts for 15% of the overall result). The relative quality (compared to other schools) of a school's student body-as judged by median LSAT scores, college GPAs, and the school's acceptance rate-also has tended, as an historical matter, not to change tremendously in a single year (but rather evolves much more gradually), but this factor has itself become a bit more volatile in recent years as the national decline in application volume has hit some schools harder than others. Other factors, such as the percentage of graduates who are placed in law-related jobs at or nine or ten months after graduation, bar pass rates, and dollars-per-student spent by a school (more on that later), have tended to fluctuate much more, and thus may account more for the year-to-year changes in bottom-line rankings.

One might argue that the parts of the U.S. News survey that are more stable are more reliable and thus should be taken more seriously than the overall rankings. There is something to that, but even these stable components have been open to significant criticism. The response rate by lawyers and judges who are polled has often been quite low, and the integer-based scale (ranging from 1 to 5) on which law professors, lawyers and judges are asked to place schools is not sufficiently finely grained for people to draw the kind of nuanced distinctions that the U.S. News rankings purport to depict overall. Moreover, it may be that larger law schools, with more graduates, may have an easier time making a positive impression on judges and lawyers, simply because members of the bench and bar may be more likely to encounter recent alums of schools that pump out more graduates. (There are other, smaller aspects of the U.S. News methodology-such as student-faculty ratios-that might tend to inadequately reward economies of scale and thus favor small schools.)

#2 The U.S. News Data Is Necessarily Limited in Scope

In addition to being limited in time, the data that U.S. News employs and presents every year is limited in scope. Among the data that it ignores is how diverse a law school's faculty or student body is. We have argued (in an earlier series of online columns) that this information concerning racial/ethnic (and perhaps other kinds) of diversity ought to be incorporated into the rankings. Most law school faculty and administrators around the country believe that diversity within a school is a helpful plus in a world where graduates are going to encounter and serve clients of various different backgrounds. Yet U.S. News has declined to include a diversity component in its overall scoring (although it separately presents raw data as to racial diversity). As we have explained before, the main reason U.S. News has offered for not including diversity-that some schools are located in places where diversity is harder to accomplish-simply doesn't wash. Some schools are located in places where there are fewer high-LSAT performers in the community, yet we still include median LSAT as a rankings input because we think a law school student body's LSAT performance is a relevant characteristic. If, as the Supreme Court has held and as most people in academia believe, a diverse school is pedagogically better than a less diverse school, all other things being equal, then we should develop a way to have diversity count for at least something when we evaluate and rate schools. In the meantime, prospective students can find helpful data on each school at the ABA "Standard 509" website.

#3. Distributions Within Each Law School Student Body

While we are talking about the makeup of the student body of each law school (which many prospective students would find an important factor since law students often learn from, and are judged by the outside world by, the company they keep), we should point out that the data that U.S. News weighs most heavily-median LSATs and GPAs-while relevant to an assessment of student-body academic strength, itself can mask important differences within each student body. Two law schools may have similar medians, but they may have very different LSAT scores and GPAs at the 75^th and 25^th percentiles within their student bodies. Let us compare, for example, using 2014 data, Northwestern and Cornell, both excellent law schools with undeniably strong student bodies. Northwestern's LSAT median was a 168, and its median college GPA was a 3.75. Cornell's were a bit lower on both-a 167 and a 3.68. But Cornell's 25^th percentile LSAT and GPA were somewhat higher than Northwestern's (166/3.55 compared to 162/3.53). How could the school with higher medians have lower numbers at the 25^th percentile? There could be a number of possible explanations. Northwestern may prefer applicants who have either a very high LSAT or a very high GPA (sometimes known as "splitters"), whereas Cornell may prefer people who were reasonably (but not quite as) high on both metrics. Perhaps Northwestern's 25^th percentile LSAT is lower because it has enrolled more students who have been out of college for a longer period of time, in which case LSAT scores may be less important than real-world accomplishment. Or maybe some different reason altogether.

We are not suggesting here that having medians that diverge from a school's 25^th percentile numbers is inherently problematic (although it might be problematic for law schools that, unlike Cornell and Northwestern, have many low LSAT performers and that may have low bar pass rates); instead, we are simply saying that when an applicant is looking at the student bodies of schools in which s/he is interested, it may make sense to look at more than medians. We note, in this regard, that the ABA "Standard 509 Report" website has a good tool that enables users to search and compare law schools along these axes.

We should add that if critics believe that U.S. News creates a perverse incentive for schools to admit "splitters" (perverse in the sense that pedagogical considerations would otherwise incline these schools to admit folks who present reasonably strong LSAT scores and GPAs instead), there might be ways to tweak the U.S. News student-quality formula, but any such changes could create other incentives or disincentives about which other observers might complain.

#4. Looking Behind the Employment Numbers

Two final observations warrant mention. First, one of the most volatile-and thus influential as to many schools' rankings in a given year-factors in U.S. News is the percentage of graduates who have a full-time, long-term, law-related job ten months after graduation. Certainly a school's ability to help place its graduates is an important factor in any decision about where to attend law school. But note, importantly, that the percentage employed in full-time, long-term, law-related jobs does not by itself convey any information about the particular type of jobs a school's graduates are getting. U.S. News does not present or make use of salary data (although it did decades ago); it does not break jobs down by geography; as of last year it did not even tell consumers how many jobs are funded by the graduate's law school or home university. We should add that some law school- or university-funded jobs are quite meaningful and reasonably paid, whereas others are less so. In any event, here too, the ABA provides much more finely grained data on job type and salary; applicants should consult the web page on which the ABA collects and presents the employment surveys for all ABA-approved schools.

#5. Follow (or at Least Examine) the Money

Finally, speaking of money, we should point out that there is one factor in U.S. News as to which the underlying data and the use to which U.S. News puts it are harder to see and thus harder to analyze, and that is the so-called "faculty resources" component that looks at "average fiscal expenditures per student for instruction, library and supporting services." This inscrutable factor accounts for about 10% of a school's overall score and often determines where a school lands within a bunched-up grouping. For example, if one looked at all the other major U.S. News components-peer academic assessment, lawyer/judges assessment, median LSATs/GPAs, acceptance rates, placement rates, bar pass rates, student-faculty ratio, etc.-Yale should be tied with, or even slightly behind Harvard. And yet Yale consistently beats Harvard for the top spot in the rankings by a non-trivial margin; last year it was four overall score points out of a possible 100. And this difference seems likely accounted for by the fact that Yale spends more-although precisely how much more is hard to know-per student than any other school by a significant margin.

Now four points out of 100 in the U.S. News overall score may not seem like a lot, but given how bunched up schools are, four points can be a big deal. (Yale's four-point lead over Harvard last year was four times larger than Harvard's lead over #3 Stanford, and four points farther down the scale was all that separated #29 from #42.) And Yale's perch atop U.S. News every single year for over two decades likely accounts for its qualitatively better yield among admitted applicants than that of any other law school, thus enabling Yale essentially to have first choice among the applicant pool. (Yale these days admits only around 250 people to get 200 to attend, generating a yield of about 80%, compared to Harvard's yield of about 60%, which itself is much higher than the yields of almost all other top law schools.)

To say that expenditures-per-student can have these important consequences on a school's ranking and its yield is not to imply that this spending criterion is illegitimate. But one cannot help wondering: if the additional spending-per-student isn't elevating placement rates or lowering student-faculty ratios, or allowing a school to obtain a faculty that is seen by other law professors as superior to that of other schools (and all of these are already measured directly and counted by the ranking system), precisely why should the money matter so much? The U.S. News methodology may result in double-counting of many considerations, but dollars spent may be a particularly problematic example. Yet there may be responses: perhaps Yale's resources don't increase its placement rate, but affect the kinds of jobs its graduates are able to obtain. For example, maybe its resources allow more students to undertake their own original research, which leads to more jobs in the academy. And so forth. Again, as with most other features of the U.S. News rankings that we've discussed above, our goal here is not so much to provide definitive answers as to cause students to think a bit more critically as they consume the bottom-line ordinal rankings for which U.S. News is best known.

Cross-posted from Justia's Verdict.

This week offered quite the spectacle in Alabama. Relying on a recent ruling from United States District Court Judge Callie Granade, based in Mobile, that struck down the provision in the Alabama state constitution that prohibits recognition of same-sex marriage, many gay and lesbian couples around the state began getting marriage licenses. But other same-sex couples, mainly in more conservative counties, have been unable to obtain licenses because some state probate judges (who issue marriage licenses in that state) are continuing to abide by the state-law ban on same-sex marriage, notwithstanding Judge Granade's ruling that such discrimination violates the Fourteenth Amendment of the federal Constitution. And on Monday Judge Granade herself declined requests to hold probate judges who refused to issue same-sex licenses in contempt of her federal court and its orders. Meanwhile, the Chief Justice of the Alabama Supreme Court, Roy S. Moore, has been issuing statements about the limits of federal power that remind many observers of the days of George Wallace, and maybe even Jefferson Davis. So what is going on here? And do the state court judges who are continuing to enforce Alabama's discriminatory marriage-license regime have a legal leg to stand on? In the space below, I try to separate the strands of the tangle, and to highlight which legal questions have clear answers and which don't.

Let us first take the easy question of whether Justice Moore's broad-based challenge to federal judicial authority holds up. It does not. Justice Moore has said that the federal courts have no authority over the state-law institution of marriage and that federal district judges cannot require state judges to follow federal trial court rulings. These ambitious sentiments are certainly wrong if they are taken to mean that a federal court cannot award meaningful relief to plaintiffs who successfully challenge the application of state law to them. Federal district courts can vindicate the federal rights of federal plaintiffs, even if those rights involve the institution of marriage, and even if it is state judges (as is the case in Alabama's regime) who issue the marriage licenses that are being wrongfully and unequally withheld. So a federal district court has undeniable power to order state officials, including state judges, to provide victorious federal plaintiffs a remedy to redress their constitutional violations. Such power to adjudicate and vindicate federal rights is emphatically what federal courts are for. That's Constitutional Law I/Marbury v. Madison-kind of stuff.

The Limits on Federal District Court Remedial Reach

Why, then, did District Court Judge Granade not hold state court judges in contempt for withholding marriage licenses? After all, contempt of court-with its coercive sanctions-is usually what we invoke to ensure that people comply with federal court orders. (I should point out here that individuals who violate federal court orders are generally punishable for contempt even when the federal court orders themselves go too far and are later determined to be illegal.) The answer is that the formal remedy provided by Judge Granade-the technical order she issued after finding the Alabama same-sex marriage ban unconstitutional-did not, as she herself understood it, extend to all probate judges who interact with all same-sex couples throughout the state. While Judge Granade could and should hold in contempt any state official who refuses to recognize the marriage of the particular same-sex couple who brought the case in her court and won, Judge Granade was likely correct not to try to punish probate judges for withholding relief as to other same-sex couples.

The reason for this is that the weight of authority tends to suggest-as a leading casebook puts things-that "a [federal district] court can enjoin [a] defendant only with respect to the defendant's treatment of plaintiffs actually before the court, either individually or as part of a certified class" (emphasis added). Because there was no class certified in the case before Judge Granade-it was brought by one same-sex couple-Judge Granade's remedial authority is technically limited to the particular plaintiffs before her. Thus, even if her legal reasoning invalidating Alabama's same-sex marriage ban is valid-and even if it is likely that her interpretation of the federal Constitution will be upheld by the U.S. Supreme Court later this year-state court judges who continue to enforce the state-law ban as to other couples are likely not defying federal authority in a way that can be punished.

This also means that, as a technical matter, the problem (if one views remedial limitations as a problem, though many would view them as a virtue that makes federal judicial power less scary) won't necessarily be solved by trying to name every probate judge in Alabama as a party to a case in Judge Granade's docket. If a federal judge cannot order state judges to provide relief to anyone other than the federal plaintiffs before her, then same-sex couples throughout the state will not all necessarily benefit by extending Judge Granade's orders to cover additional defendants in additional cases involving additional named couples; the remedial limitation is defined by the identity of the plaintiffs, not the identity of the defendants.

So when Judge Granade yesterday afternoon ruled, in a second case, that one probate judge who had been refusing to issue same-sex marriage licenses must begin issuing them, her order was still limited to the particular plaintiffs (four same-sex couples) who were in her court asking for relief. As Judge Granade wrote yesterday: "If Plaintiffs take all steps that are required in the normal course of business as a prerequisite to issuing a marriage license to opposite-sex couples, Judge Davis may not deny them a license on the ground that Plaintiffs constitute same-sex couples or because it is prohibited by the Sanctity of Marriage Amendment and the Alabama Marriage Protection Act or by any other Alabama law or Order pertaining to same-sex marriage" (emphasis added).

Now it may well be that as more and more probate judges are instructed to issue licenses to named plaintiffs in more and more cases, all of the probate judges in the state will start issuing licenses to all same-sex couples who apply, regardless of whether those couples are victorious plaintiffs in any federal case. But that will not be because the state court judges are under a federal court order that obliges them to do so, but because they choose to do so in light of the decisional trend.

I should note that the conventional rule that I describe here-that a federal trial court's relief should normally be granted only to the actual plaintiffs in the case-does not forbid the court, even outside of a class action setting, from ordering relief that in fact goes beyond protecting the named plaintiffs and also protects other would-be plaintiffs, if full relief cannot be given to the named plaintiffs without also necessarily regulating the defendants' interactions with other persons. (For example, in one case, a court was justified in ordering the police to stop enforcing a motorcycle helmet law overly aggressively as to all riders-and not just as to the plaintiffs-because highway patrol officers would have no way of distinguishing plaintiffs from non-plaintiffs before pulling someone over.) But in the present setting, full relief (i.e., marriage licenses and recognition) can be given to named plaintiffs without ordering the defendants to give licenses to anyone else.

Some have argued that providing full relief to named same-sex plaintiff couples requires allowing all same-sex couples in the state to marry, because absent such broad access to same-sex marriage, the named plaintiffs' marriages would continue to be subject to stigma. But I don't think that this stigma argument works, because if it did, then same-sex couples whose marriages are already recognized would have standing to challenge bans on same-sex marriage that are preventing others from marrying, and I don't think any federal court would recognize standing in such circumstances.

I should also point out that some federal judges believe that a district court can order government agencies to refrain from enforcing facially invalid laws or policies against anyone, and not just the plaintiffs before it. (Judge Granade's orders up until this point-limited as they are to the actual plaintiffs before her-give no indication that she is among them.) For example, (now-retired) United States District Court Judge Vaughn Walker (in San Francisco) is reported to have embraced this view when he issued an order whose plain language directed state officials to stop enforcing California's same-sex marriage ban, Proposition 8, against all same-sex couples, and not just the two couples who sued in his court. Judge Walker's apparent position-which was never fully tested because both the Governor and Attorney General of California chose not to try to continue enforcing Proposition 8-was in (unexplained) tension with current Ninth Circuit law, which embraces the more dominant view, described above, to the effect that the remedy must ordinarily be tailored to the plaintiffs only. The Ninth Circuit approach is supported by most (but not all) of the statements the Supreme Court has made on the topic, but candor compels the acknowledgement that the law in this area is not really settled and could definitely benefit from high Court attention and clarification.

Does Restricting a Federal District Court's Reach to the Plaintiffs Before it Make Sense?

Why might it be sensible for a federal district court judge not to be able to issue relief to anyone other than the plaintiffs in the case before it? Because, under the judicial system we have chosen, we have decided that federal district court opinions and decisions should, as a matter of governing precedent, have no binding effect on any other judges, even other federal judges located within the same district. This situation is to be contrasted with a ruling by the regional federal Court of Appeals or the United States Supreme Court. Once either of those courts has held Alabama's law invalid (and neither has yet-the Supreme Court chose not to block Judge Granade's ruling, but it won't decide the merits of the same-sex marriage constitutional question until later this year, at the earliest), then all judges, state and federal, within the state should surely obey that ruling, because the federal appellate court (whether it is the Eleventh Circuit or the Supreme Court) would have fashioned federal law that is supreme and applicable throughout the state. This is true even though state judges' rulings are not appealable to the Eleventh Circuit, insofar as all federal district judges in the circuit (who are bound by circuit precedent) would have no choice but to give injunctive relief to any same-sex couple who subsequently filed suit. Under those circumstances, it would be an utter waste of time (and perhaps a due process violation) for a state court judge not to give a license to someone who undeniably could get one by filing a federal suit anywhere in the state. Whether contempt sanctions are applicable or not, no judge or other state official would be justified in continuing to enforce a state law that a federal appellate court governing that state has held to be invalid.

But a ruling by a district court judge like Judge Granade has no such effect. Just because she ruled that Alabama's law violates the federal Constitution does not mean that other federal judges in Alabama would so rule if other same-sex couples filed suit in their courts. Her ruling is not binding precedent on them. Importantly, not all same-sex couples could properly sue in Judge Granade's district, and even if they did, other district court judges in that district to whom a new case might be assigned might rule differently on Alabama's ban on same-sex marriage. So Judge Granade's ruling-unlike one from the Eleventh Circuit or the Supreme Court-does not inevitably provide relief to any would-be federal-plaintiff same-sex couple in the state.

Perhaps an example will help drive home the point I'm making. Imagine that public universities throughout Alabama, pursuant to a state law policy, take race of applicants into account in a measured way in the admissions process, in order to assemble a diverse student body. Suppose a single unsuccessful applicant to a single public college in the state sues in federal court, bringing a facial challenge to the state's affirmative action policy on the ground that any use of race violates the Fourteenth Amendment. And suppose the district judge in that case rules (wrongly, to my mind, but not implausibly as a prediction of where the Supreme Court is headed) that all consideration of race is indeed barred by the Fourteenth Amendment. Could that judge apply her ruling to all the public universities in the state, and order all of the them (under pain of contempt) to refrain from considering race at all as they process the hundreds of of thousands of applications they receive each year, even though many other federal and state judges in the state would disagree with her interpretation of the Fourteenth Amendment but might never have a chance to hear a case if her ruling were allowed to shut down statewide use of the policy? I think such remedial power by a single judge would raise many problems, and for that reason, if and when the Court clarifies district court remedial power, it might embrace what I have described here as the mildly dominant view limiting remedial authority to actual plaintiffs.

But, a reader might wonder, isn't there a difference between the affirmative-action hypothetical I posit and the same-sex marriage setting insofar as affirmative-action limitations at the high Court are still a matter of debate, while there is no longer any real doubt about whether the Supreme Court this summer will hold that all state-law discrimination against same-sex couples in the marriage arena is unconstitutional? Perhaps this prediction is quite sound, and state court judges would be justified if they chose to issue licenses on that basis, but I am not entirely sure that state court judges are required, as a matter of supreme federal law, to act on it now. All judges have a legal obligation to follow binding precedent from a higher court once it is handed down, but there may be no legal obligation-enforceable by contempt or otherwise-on lower court judges to see the writing on the wall.

Cross-posted from Justia's Verdict.

In my last column, I explored some aspects of an important case, Arizona Legislature v. Arizona Independent Redistricting Commission, pending at the Supreme Court. As I explained, the merits question presented is whether the people of a state may create an independent redistricting commission (IRC)-i.e., one that is not controllable by the elected state legislature-to devise congressional districts, as Arizona voters did in 2000. The elected Arizona legislature (acting as a body) brought suit, arguing the so-called Elections Clause of Article I of the Constitution (Article I, section 4)-which provides that "[t]he [districts for] Representatives . . . shall be prescribed in each State by the Legislature thereof; but the Congress may at any time by Law make or alter such Regulations"-protects elected state legislatures from state laws that take congressional districting out of their hands. In the space below, I continue to explore the questions the case raises, especially in light of the additional briefs that have been filed.

Does the Elected Legislature Have Standing to Sue the State (Voters)? The Two Key Precedents

One issue Court will take up-indeed, an issue the parties initially did not address but on which the Court specifically sought briefing-is whether the elected legislature has "standing" to challenge the Arizona initiative in federal court. Elected state legislatures have been found to have standing in a number of cases in federal court, but the more ordinary situation in which an elected legislature seeks to be in court involves the legislature's attempt to defend rather than attack state law. When the executive branch of a state does not defend a state law that is challenged by private individuals, the elected legislature may under certain circumstances be permitted to do so instead.

In the Arizona case, by contrast, the elected legislature seeks to invalidate, not preserve, the Arizona law that voters passed in 2000. To do so, the legislature, like any plaintiff in federal court, must demonstrate that it has suffered (or is reasonably certain to suffer) an "injury" that is "cognizable." The elected legislature's asserted injury here is that it has been removed from an important job that the federal Constitution (in the Elections Clause) assigns directly to it. Because, under the Arizona initiative, any congressional districting legislation passed by the elected legislature will not be put into effect, the elected legislature's vote on any such districting will be (improperly, to its way of thinking) nullified.

There is one older Supreme Court case that may support the Arizona legislature's standing argument. In Coleman v. Miller (1939), a majority of the Kansas state senate brought suit to challenge the actions of the state executive branch in connection with the ratification of a proposed federal constitutional amendment dealing with child labor. The state senate had deadlocked 20-20 on the question of ratification of the amendment, an outcome that ordinarily would be construed as a decision not to ratify. But the lieutenant governor of the state (as presiding officer of the senate), decided to cast a vote-as he would in ordinary legislation-and voted in favor of ratification.

When state officials prepared to communicate that Kansas had ratified, for purposes of determining whether three-quarters of the states had ratified (the threshold required for an amendment to go into effect), the state senators who had voted against ratification, joined by three others to make a majority of the senate, sued, claiming that the lieutenant governor had no business participating in the ratification vote because Article V's conferral of power to state "legislatures" to ratify federal amendments excludes participation of state executive officials. As a result, the Kansas senators argued, their decision not to ratify (by an equally divided vote) was being improperly overridden.

The U.S. Supreme Court ruled 5-4 that the plaintiffs had standing, reasoning that the lieutenant governor's actions, if indeed violative of Article V, completely and improperly nullified the valid votes of the elected state legislators. Because their votes had been unconstitutionally ignored altogether, they had suffered an injury cognizable in federal court.

Coleman was explained, distinguished and perhaps narrowed in 1997 in Raines v. Byrd. In that case, a handful of U.S. Senators and House members brought suit to challenge the constitutionality of the federal Line Item Veto Act (LIVA), a statute passed by Congress and signed into law that purported to give the President, with respect to each future budget bill in which Congress had not indicated an intent otherwise, the power to sign the budget bill into law but then decline to spend any money on certain budget items of his choosing. The plaintiffs in Raines claimed that giving the President such authority diluted the power of Congress, because any votes on subsequent budget bills in Congress might not be given full effect by a President who decided to spend on some, but not on all, the budget items Congress had adopted.

The Court in Raines found plaintiffs lacked standing. It distinguished Coleman in at least three ways: (1) in Coleman, a majority of the Kansas senate voted to sue, whereas Raines involved only a handful of members of Congress, and neither house of Congress authorized the suit; (2) in Coleman, the vote of the twenty Kansas senators was being nullified altogether by the lieutenant governor's actions, whereas in Raines the dilution or diminution of the "effectiveness" of Congress's votes in any future budget bill may not have been as extreme; and (3) the Kansas state senate had already voted on the ratification measure in question in Coleman (and the effect of its vote in a particular case was thus at stake), whereas in Raines the alleged harm related to future votes Congress might cast.

How Should Coleman and Raines Play Out in the Arizona Case?

In some ways, the Arizona case is similar to aspects of both Coleman and Raines, a feature which gives the Court some leeway to resolve the standing question any way it wants without having to formally overrule a past case. Like Coleman, the Arizona case involves a suit by a majority of a legislative branch-and not just a few individual members. Also similar to that in Coleman, the claim here is not just that the (redistricting) laws by the elected legislature might be affected or influenced by the Arizona initiative, but that they are foreclosed altogether: all the votes by the elected legislature on a districting bill would be completely ignored. But as in Raines, the alleged harm to voting power is in the future, insofar as the elected Arizona legislature has not actually cast any redistricting votes that have been (or are about to be) ignored or nullified on account of the Arizona initiative.

Here's another potentially relevant factor. In Coleman, the injury to the Kansas senate was inflicted from outside the senate, by the lieutenant governor. The same is true in Arizona, insofar as the People wrested power from the elected legislature without the legislature's consent. In Raines, by contrast, Congress itself passed the LIVA that some of its members believe improperly diluted Congress's own power. The Raines Court did not actually rely on the self-inflicted character of the alleged institutional injury, but that may be a significant background fact.

And here's yet one more possible consideration. In Raines, the Court suggested that even if the members of Congress lacked standing to challenge the LIVA, someone else outside Congress-the intended beneficiary of a spending item that Congress approved but that the President cancelled under the LIVA-would be able to sue later to challenge the Act. And, in fact, such a challenge did occur (and the LIVA was struck down-wrongly, to my mind) in Clinton v. New York. In the present case, it is possible that a voter or congressional candidate could sue to challenge the Arizona initiative, claiming that the district in which she finds herself on account of the lines drawn by the IRC is less desirable to her than the district in which she would have been located had the elected legislature retained control, but it is far from clear that such a case would actually be filed and survive the standing hurdle. The Supreme Court has elsewhere said that just because it is hard to imagine anyone other than the plaintiff before it who would have a better claim of standing is no reason to relax standing rules, but the presence or absence of better plaintiffs might be an unstated factor in a very flexible standing doctrine. (There are intimations of that in Raines itself.) If the Arizona legislature is correct that the federal Constitution gives it particular power that is being wrongly taken away from it, a sensible system should allow someone to go to court to fix the constitutional violation.

In the end, I think the Court can-and could very well-go either way on the standing question. I note that if the Court limits Raines and allows the elected Arizona legislature to sue, it might be open to the criticism that it manipulates standing rules out of a perceived hostility to direct democracy. Two years ago, in Hollingsworth v. Perry (one of the same-sex marriage cases), the Court used questionable reasoning (even if its result was correct) to make it hard for proponents of initiatives to defend those initiatives in federal court when elected state officials decline to defend. If the Court in the pending Arizona case relaxes the standing bar to make it easier for the elected legislature to attack the Arizona initiative, some will think the Court is just plain anti-initiative.

Back to the Merits and the Key Question of Congressional Approval Power

Of course, one way to avoid that perception would be uphold the Arizona initiative on the merits. As I argued in my last column, I think there is a compelling argument on the merits that Congress, in 2 U.S.C. § 2a(c), approved the use of initiative and other direct democracy devices in the drawing of congressional districts by states, and that such a decision by Congress should be controlling regardless of whether the word "legislature" in the Elections Clause means elected legislature only or something else. The more I have examined the congressional statute in question, the more controlling I think it is, because its text is quite broad and clear in allowing states to use whatever state law devices they want to conduct districting, and because the legislative history suggests that one reason Congress wanted states to be able to use direct democracy in this arena was that elected legislatures were prone to engage in mischievous gerrymandering, the very problem to which the Arizona initiative was directed. So I think the congressional statute here is right on point.

The key question - and one that the briefs don't fully engage - then becomes whether Congress has the power to authorize states to use initiative devices to draw district lines. The Arizona elected legislature says no, but as I pointed out in my last column, the Supreme Court (in Ohio ex. rel. Davis v. Hildebrant), in upholding Ohio's use of the referendum in districting), relied explicitly on Congress's having, in adopting (the predecessor to) 2 U.S.C. § 2a(c), invoked its Article I, section 4 powers, which "expressly gave [Congress] the right to" act in this realm. As I observed, Congress, in exercising its power, might have passed a law creating the very identical Arizona IRC to do the districting within the state, and that would be completely permissible. If Congress could have enacted the IRC law itself (or incorporated it by reference into binding federal law shortly after the Arizona voters approved it), then why can't it simply approve, before the fact, any districting approach it wants? The recently filed brief for the IRC points out that it is much more convenient for Congress to approve state laws prospectively than it would be to monitor what states are doing and then enact laws itself. But that doesn't quite answer the question whether Congress has the authority to approve state laws that aren't on the books at the time Congress adopts the approval. The IRC's brief, which is superb overall, doesn't delve deeply into this matter, and the federal government's amicus brief is the only one I've seen that has more engagement with this question.

The best (albeit losing) argument against such congressional power is that prospective approval is an impermissible delegation of congressional authority to states. In the nineteenth century, such an argument might have had traction. Chief Justice Marshall in the well-known case of Gibbons v. Ogden opined that Congress cannot enable states to legislate when the Constitution disabled them from doing so because such prospective empowerment would in effect constitute a delegation of federal legislative authority back to the states. And as Justice Story observed in 1838, federal statutes that approved or incorporated state laws were generally construed as approving or incorporating state laws in effect at the time Congress acted, because there are "very serious doubts, whether [C]ongress does possess a constitutional authority to adopt prospectively state legislation on any given subject; for that, it seems to me, would amount to a delegation of its own legislative power."

But all this changed in the 1900s. In two seminal cases, the Court signaled that prospective incorporation of state laws by Congress, or prospective congressional approval of state laws that would otherwise violate the Constitution, is allowed. In United States v. Sharpnack (cited by the United States in its amicus brief), the Court allowed Congress to incorporate state criminal laws for use as federal laws in federal enclave (donut hole) territories, and the Court did not construe the incorporation as static, but instead as ongoing, incorporating into the federal law state laws that were passed after Congress acted. In rejecting a delegation attack, the Court said that rather than being a delegation by Congress of its legislative authority to the states, "[the 1948 Act] is deliberate continuing adoption by Congress for federal enclaves of such . . . offenses and punishments as shall have been already put in effect by the respective states for their own government. Congress retains the power to exclude a particular state law from the assimilative effect of the Act." Thus, the prospective adoption does not constitute a delegation because Congress remains free to withdraw the power being exercised by the states if Congress disapproves. The opportunity to reclaim the delegated authority, under the Court's reasoning, dissolves the delegation issue.

An even more important case, one I haven't seen anyone cite in the Arizona case briefs, deals directly with congressional approval of state laws that would otherwise violate the Constitution. (I tend to think the congressional statute authorizing direct democracy in drawing district lines, 2 U.S.C. § 2a(c), as more of an authorization of state law than as an incorporation of state laws into federal law, since I don't think the ins and outs of the Arizona initiative are themselves federal law.) In Prudential Ins. Co. v. Benjamin, decided in 1946, the Court effectively held when the Constitution deprives states but not Congress of authority to do certain things, it does not restrict the "coordinated exercise" of federal and state authority. Put another way, if Congress can do something alone, Congress can consent (oven prospectively) to having the states do it instead. As Professor Cohen has correctly observed, the Court's theory sweeps broadly: "Congress may remove all constitutional limits on States when those limits are wholly inapplicable to Congress-that is, when they stem solely from divisions of power within the federal system."

As I have explained more fully in academic writings, I think the twentieth century attitude reflected in Sharpnack and Prudential may have something to do with the fact that, beginning in the early 1900s, U.S. Senators were no longer elected by state legislatures, such that delegations by Congress to state governments were less scary, insofar as states (through their clout over Senators) wouldn't be able to block efforts by Congress to reclaim federal power if states were abusing it. As Sharpnack pointed out, as long as Congress can pull back any power it has given to states, the delegation problem is minimized.

Perhaps delegation to state peoples to engage in direct democracy, as opposed to delegations to elected state legislatures, never raised reclamation problems even before the 1900s, so that even Chief Justice Marshall and Justice Story, in their times, would see no problem with a federal law that allowed states, freely and prospectively, to make use of direct democracy in congressional district drawing. But in any event, in light of Sharpnack and especially Prudential, the congressional statute at issue in the Arizona case, 2 U.S.C. § 2a(c), is a permissible exercise of congressional power, and thus should be an easy basis on which the Court could resolve the case, if it chooses to reach the merits at all.