The Court returned this morning, announcing decisions in two of its most-watched cases of the term. In Our Lady of Guadalupe School v. Morrisey (No. 19-267), a seven-justice majority held that the First Amendment bars courts from hearing employment discrimination claims brought by a teacher in a religious school. And in Little Sisters of the Poor Saints Peter and Paul Home v. Pennsylvania (No. 19-431), the Court upheld two administrative rules issued by the Trump administration allowing employers who object to providing their employees with health insurance coverage for contraceptive services to opt out of the ACA’s contraceptive mandates on religious or moral grounds.

But that’s not all. Just after it wrapped up issuing the decisions in these cases, the Court announced that tomorrow it will issue its final decisions of the term. What’s still to come? McGirt v. Oklahoma (No. 18-9526), for one, an exceedingly strange case that has the potential to significantly change the law regarding states’ authority to prosecute crimes committed in Indian country. And three cases—Trump v. Vance (No. 19-635), Trump v. Mazars USA, LLP (No. 19-715), and Trump v. Deutsche Bank AG (No. 19-760)—addressing the validity of subpoenas issued to Trump-affiliated entities by state prosecutors and the House of Representatives seeking Trump’s tax returns and related documents. We’ll let you know how all that turns out. But in the meantime, we’ve got summaries of the Court’s two decisions from Monday: Chiafalo v. Washington (No. 19-465) and Barr v. American Association of Political Consultants (No. 19-631).

As the 2000 and 2016 presidential elections showed (and for the history buff among us, the 1824, 1876, and 1888 elections, as well), American voters don’t directly elect the President. Instead, their votes go toward selecting members of the Electoral College. Each member of a state’s delegation to that prestigious group is appointed by her state in accordance with that state’s popular vote returns. Most states require electors to pledge that they will support the nominee of the party whose candidate won the state’s popular vote. In Ray v. Blair (1952), the Supreme Court affirmed the constitutionality of these pledge requirements. Some states go one step further, however, by imposing fines or other penalties on electors who violate their pledge. In Chiafolo v. Washington (19-465) (and a companion case, Colorado Department of State v. Baca (No. 19-518)) the Court was asked whether such enforcement mechanisms are permitted under the Constitution. Justice Kagan, writing for a nearly unanimous Court, answered in the affirmative. Only Justice Thomas did not join the opinion, writing separately to concur in the judgment.

Article II § 1 of the Constitution, as amended by the Twelfth Amendment, provides for the Electoral College, as we know and love it/put up with. The text is sparse on details, providing only a general directive to states to appoint electors and to ensure that the electors vote separately for the president and vice president. By the twentieth century, most states had enacted statutes meant to prohibit so-called “faithless voting,” which require electors to vote in line with their state’s popular vote. Fifteen states back up those statutes with some form of sanction, such as imposing a fine or removing the faithless elector from his or her position.

At the time of the events in question, Washington was one such state: Electors had to pledge to vote in accordance with the party nominating them (i.e., the party that won Washington’s popular vote) and could face a fine of $1,000 for violating the pledge. (Washington has since rescinded the fine.) Three Washington electors violated their pledge and cast ballots in favor of Colin Powell rather than Hillary Clinton, hoping to convince electors in red states to act similarly and thus deprive Donald Trump of a majority in the Electoral College. After the three were each fined $1,000, they challenged the fine as unconstitutional. The Washington Supreme Court affirmed the fine. Similar shenanigans occurred in Colorado, where pursuant to that state’s law, Michael Baca was removed from his position as an elector when he tried to vote for John Kasich. But unlike the Washington Supreme Court, the Tenth Circuit held that Colorado’s faithless elector law was unconstitutional. The Supreme Court granted cert to resolve the split, and then sided with the Washington Supreme Court, affirming that court’s judgment and reversing the judgment of the Tenth Circuit.

After laying out this procedural history and context, Justice Kagan began her substantive analysis by noting that the Constitution gives states “far-reaching authority over presidential electors, absent some other constitutional constraint.” This power carries with it the power to impose conditions on an elector’s appointment, including, for example, that the elector live in the state. Given the breadth of states’ authority over electors, there is nothing preventing a state from also enacting enforcement mechanisms to ensure electors do not cast faithless votes. Justice Kagan then went on to note that, had the Framers intended to restrict states from making such laws, they could have done so—indeed, at the time of the founding, both Maryland and Kentucky required electors voting for state senators to swear an oath that emphasized each elector’s “judgment and conscience.” According to Justice Kagan, if such language had been included in the Constitution, then electors might have more freedom. But no such language appears, so states are free to restrict electors’ discretion as they wish. The three Washington electors argued that the very words “electors,” “vote,” and “ballot,” which appear in Article II § 1 and the Twelfth Amendment, necessarily connote an exercise of judgment and discretion. The Court rejected that argument, noting that someone who “always votes in the way his spouse, or pastor, or union tells him to” is still “voting” even though “the choice is in someone else’s hands.” The Court also noted that in some elections there is only one candidate, yet if someone “goes through the motions” of checking the only box available, we would still ordinarily say he has “voted.”

Having thus concluded that the Constitution’s text does not bar states from enforcing pledges to avoid faithless voting, Justice Kagan turned to history. At the very first contested election in 1796, every elector but one voted in accordance with their pledge. Throughout the nineteenth century, courts and commentators alike emphasized that electors were meant only “to register their votes, which are already pledged,” and that an “independent judgment would be treated as a political usurpation.” And in practice, faithless voting has been very rare: out of the over 23,000 votes cast by electors in our history, just 180 have been faithless votes—more than a third of which were cast in 1872 when the Democratic nominee died just after election day. The Court thus concluded that both the Constitution’s text and the history of actual practice support the notion that states may require electors to vote in accordance with the popular vote, and that they may enforce those requirements with the threat of sanction.

Justice Thomas concurred with the judgement but disagreed with the Court’s reasoning. In his view, Article II and the Twelfth Amendment are silent on the question of whether states may bind electors to vote in accord with the popular vote. He would have preferred to affirm relying on on the Tenth Amendment, because “[w]hen the Constitution is silent, authority resides with the States or the people.”

While Americans may disagree about the merits of the electoral college, one thing we can all agree on is that we hate robocalls. So nearly everyone can celebrate a win in Barr v. American Association of Political Consultants, Inc. (19-631), where the Court upheld (and actually broadened) a ban on certain types of robocalls to cellphones. Who isn’t cheering? Well, the political consultants who wanted to be able to make such calls and challenged the ban. (A loss for political consultants? That’s win number two!) But here’s the strange thing: the political consultants actually won the case, successfully arguing that a portion of the statute banning some robocalls violated the First Amendment. So how did they win on the merits but still lose? Because the important part of this case is less what it has to say about the First Amendment and more what it has to say about the doctrine of severability.  

Thanks to the Telephone Consumer Protection Act of 1991 (you might know it as the TCPA), almost all robocalls to cellphones are banned. Even back then, Congress realized that they’re “the scourge of modern civilization.” But in 2015, Congress adopted an exception to this ban, allowing robocalls to be made for the purpose of collecting debts owed to or guaranteed by the federal government. That sounds like a restriction on speech based on content: If you get a robocall asking you to pay off a government debt, it’s allowed. If you get one asking for anything else, it’s not. So the American Association of Political Consultants, who want to be able to make robocalls for purpose of polling and the like, brought a declaratory judgment action alleging that the TCPA’s robocall provisions were unconstitutional. The Fourth Circuit eventually agreed, concluding that the government-debt exception was unconstitutional. But rather than striking down the entire statute, it invalidated just the government-debt exception, essentially striking it from the statute. Thus all robocalls to cell phones—whether made to collect government debts or otherwise—remained illegal. The government then sought certiorari, as it does in most cases where a court holds a federal statute unconstitutional. But unusually, the political consultants sought cert too, arguing that the right remedy was to invalidate the entire robocall ban. The Court granted cert.

The Court affirmed. Justice Kavanaugh, writing for a plurality, Justice Kavanaugh summarized the outcome in this helpful way: Six members of the Court (all but Justices Ginsburg, Breyer, and Kagan) concluded that the government-debt provision violated the First Amendment. But seven members of the Court (all but Justices Thomas and Gorsuch) concluded that the proper remedy was to invalidate just the government-debt exception and sever it from the rest of the statute. Thus, just as in the Fourth Circuit, the government-debt exception was eliminated, but the robocall ban remained in force, with an even greater scope than it had before the political consultants’ successful suit.

To the plurality, the merits of the constitutional question were pretty simple. A restriction permitting speech about some topics (government debt) but not others (well, anything else) was “about as content-based as it gets.” Content-based speech restrictions like are analyzed under strict scrutiny. And the government didn’t even try to justify the government-debt exception under that standard. QED, at least as far as the merits.

Justice Kavanaugh’s plurality decision then turned to how to fix the problem. As noted, the political consultants thought the Court should go further and declare the whole restriction on cellphone robocalls unconstitutional. Among other points, they argued that Congress’s 2015 creation of the government-debt exception showed that attitudes around consumer privacy had changed and that Congress no longer cared about the ban. Hence, if faced with a choice between a TCPA without the government-debt exception or no TCPA robocall ban at all, Congress would (they claimed) choose the latter. But the Court disagreed. While this argument might work in a statute riddled with exceptions, that did not describe the TCPA. Moreover, the Communications Act of 1934, which the TCPA was technically just a part of, included a severability clause, which provided that if any specific provisions of the act were found unconstitutional, it should be severed. Finally, and perhaps most important of all, the plurality noted the existence of a “strong presumption” that courts should sever unconstitutional pieces of laws, as opposed to invalidating the entire statutory scheme, whenever that was possible. Thus it would take strong evidence indeed for the Court to conclude that Congress would’ve preferred the entire statute to be invalidated, evidence that was lacking here.

As you probably would’ve guessed from Justice Kavanaugh’s summary of the ultimate result, there were several separate opinions. Justice Breyer, joined by Justices Ginsburg and Kagan, partially dissented, arguing that strict scrutiny shouldn’t apply. Even though the restriction was content-based, he wrote, not all content-based restrictions are equal. This one, for instance, was really just a “[r]egulation of debt collection” that has “nothing to do with the free marketplace of ideas” that strict scrutiny is designed to protect. Instead, these dissenters would have applied intermediate scrutiny, and they thought the statute survived under that standard. Regardless, they agreed with Justice Kavanaugh’s plurality decision that the exception should be severed.

Justice Sotomayor concurred in the judgment all around. She agreed with the observation of Justice Breyer’s dissent that not all content-based speech restrictions warrant strict scrutiny. But unlike the primary dissent, she concluded that the government-debt exception could not survive even that lower level of scrutiny.

Finally, we come to the concurrence/dissent of Justice Gorsuch, joined in part by Justice Thomas. He agreed that strict scrutiny applied and that the government-debt exception failed under that standard. But (joined by Justice Thomas) he questioned the validity of severability doctrine as a whole and the Court’s application of it here. First, he found it odd that the Court would could cure an unconstitutional speech restrictions by leveling others “down” to the plaintiffs’ status, rather than leveling the plaintiffs “up” to where they could enjoy maximum freedom of speech. On a more practical level, he was concerned with the effects of making robocalls illegal after people had built businesses in reliance on the statute’s passage just five years earlier. He also lamented the irony of the plaintiffs’ “relief,” where they’d won the battle (convincing the Court the exception was unconstitutional), but lost the war (having the court sever the exception and ban robocalls altogether). Rather than engaging a severability analysis at all, Justice Gorsuch agreed with an approach to severability that Justice Thomas has recently advocated for in other severability cases: simply leave the exception in place but grant the plaintiffs an injunction preventing the law’s enforcement against them.