However troubling its political implications might be, the Colorado Supreme Court’s ruling on Tuesday that Donald Trump is disqualified from the state’s primary ballot for having “engaged in insurrection” demonstrates that the judicial system is still functioning in the United States. The reason is straightforward: The court applied the plain language of the Constitution, doing its job with clarity and fidelity to the rule of law.

But perhaps what is most striking about Colorado’s decision was the conservative reasoning the justices employed to reach their conclusion. The four justices who voted in the majority adhered to three stalwart principles of judicial conservatism: textualism (by which judges endeavor to strictly apply the plain text of the Constitution), originalism (by which they refer to historical sources for a contemporaneous understanding of that text), and federalism (by which judges take pains to respect the dual sovereignty of the states alongside the federal government as well as the state courts’ concomitant prerogative to construe their own laws).

This third element is perhaps the most interesting. The Colorado Supreme Court was tasked with interpreting Colorado’s Uniform Election Code of 1992, which contains that state’s criteria for getting on its presidential ballot. It determined that disqualification under Section 3 of the Fourteenth Amendment is also disqualifying under Colorado law. And it upheld the lower court’s conclusion, after a multiday evidentiary hearing, that Donald Trump in fact engaged in insurrection. Because he is thus disqualified as a matter of Colorado law, the Colorado Supreme Court determined, “it would be a wrongful act under the Election Code for the Secretary [of State] to list him as a candidate on the presidential primary ballot.”

[George T. Conway III: The Colorado ruling changed my mind]

The U.S. Supreme Court has ignored this sort of reasoning before—and to ill effect. In Bush v. Gore, it ruled in 2000 that manual recounts under Florida’s law regarding contested election results would violate the Constitution’s equal-protection clause, and thus effectively handed the election to George W. Bush by a margin of 537 votes. In dissent, Justice John Paul Stevens emphasized that “when questions arise about the meaning of state laws, including election laws, it is our settled practice to accept the opinions of the highest courts of the States as providing the final answers.” The conservative justices of today’s Court should bear this example in mind—and the stakes for the Court’s legitimacy—when considering whether the Colorado court got this aspect of its interpretation right.

Next, consider the plain language of the Fourteenth Amendment, which belies a handful of textual ambiguities: What is “insurrection” (and did January 6 qualify)? What does “engaged” mean (and did Trump do it)? And is the president of the United States an “officer” of the United States covered by Section 3? As for the first two questions, President Joe Biden summed things up yesterday, calling it “self-evident” that Trump “supported” an insurrection. Nobody seriously contends otherwise. The rebuttal instead is that Section 3 kicks in only if a jury makes these findings beyond a reasonable doubt pursuant to a federal statute that criminalizes insurrections (and which Special Counsel Jack Smith declined to invoke in indicting Trump)—an argument that one of the dissenting justices made as well.

The Colorado Supreme Court elegantly dispensed with that concern. Again, it applied a plain reading of the law, concluding that Congress’s decision to criminalize “the same conduct that is disqualifying under Section Three … cannot be read to mean that only those charged and convicted of violating the law are constitutionally disqualified from holding office without assuming a great deal of meaning not present in the text or the law.” Neither the Constitution nor the statute say anything of the sort. The court thus refused to go where it needn’t by theorizing about inferences buried beneath the plain text, which is precisely how conservative judging, at least in theory, is supposed to work.

On the officer question, the Colorado Supreme Court focused on the Constitution as written, noting that it “refers to the Presidency as an ‘Office’ twenty-five times,” including in connection with the natural-born-citizen eligibility requirement for the presidency (Article II, Section 5), the four-year cap on presidential terms in office (also in Article II, Section 5), and the impeachment clause (Article I, Section 3). It then turned to tools of originalism, observing, for example, that “dictionaries from the time of the Fourteenth Amendment’s ratification define ‘office’ as a ‘particular duty, charge or trust conferred by public authority, and for a public purpose,’ that is ‘undertaken by … authority from government or those who administer it.’” The court then reasonably concluded that “the Presidency falls comfortably within these definitions.” Judges make these kinds of interpretative decisions all the time.

Serious constitutional scholars have nonetheless pushed back on the notion that Section 3 applies to presidents, underscoring that prior drafts of Section 3 included references to “the office of the President” but that the language was ultimately abandoned. According to this argument, the framers of Section 3 intended only to prevent insurrectionists from serving in the Electoral College, but left qualified electors free to choose insurrectionists for the presidency. But those distinctions are missing from the actual text. As the conservative scholars William Baude and Michael Stokes Paulsen argued in an exhaustive article, “The substantive terms of Section Three’s prohibition are not themselves difficult or inscrutable.” Even more to the point: Jurists differ over what tools of constitutional interpretation are paramount in construing arcane constitutional terms. The political right, for example, has long assailed progressive judges for emphasizing the purposes behind a law when a plain-text reading would arguably suffice. For conservative justices to abandon that hierarchy now, on a case this consequential, would destroy whatever guise of impartiality the Court has left.

[David Frum: The Colorado Supreme Court just gave Republicans a chance to save themselves]

If the U.S. Supreme Court winds up leaving the Colorado Supreme Court’s decision undisturbed, it will inevitably get GOP voters and politicians very upset with the justices in the majority. It could also encourage states to play fast and loose with Section 3 to keep legitimate candidates off future ballots. But the threat of political retribution is just the sort of possibility that motivated the Framers of the original Constitution to give federal judges lifetime appointments under Article III—they needn’t think about the popularity of their decisions. Moreover, the ostensible point of the so-called conservative judicial philosophies of textualism, originalism, and federalism is to confine judges to the business of judging. That means resolving, on the narrowest possible grounds, discrete disputes affecting the immediate parties, at least one of whom is concretely injured by the other—rather than wading into political or normative policy conundrums in ways that aggrandize their own power relative to that of the other branches of government. If the purportedly conservative members of the U.S. Supreme Court are intellectually honest about their jurisprudential approach to the law, this case should not be hard.

When I review divided appellate-court decisions, I almost always read the dissenting opinions first. The habit formed back when I was a young law student and lawyer—and Federalist Society member—in the late 1980s, when I would pore (and, I confess, usually coo) over Justice Antonin Scalia’s latest dissents.

I came to adopt the practice not just for newsworthy rulings that I disagreed with, but for decisions I agreed with, including even obscure cases in the areas of business law I practiced. Dissents are generally shorter, and almost always more fun to read, than majority opinions; judges usually feel freer to express themselves when writing separately. But dissents are also intellectually useful: If there’s a weakness in the majority’s argument, an able judge will expose it, sometimes brutally, and she may make you change your mind, or at least be less dismissive of her position, even when you disagree. Give me a pile of Justice Elena Kagan’s dissents to read anytime—I love them even when she’s wrong, as I think she often is. You can learn a lot from dissents.

Last night, I reviewed the three separate dissents in Anderson v. Griswold, the landmark 4–3 Colorado Supreme Court case holding that Section 3 of the Fourteenth Amendment prohibits Donald Trump from ever serving again as president of the United States. I had been skeptical of the argument, but not for any concrete legal reason. To the contrary, I believed the masterful article written by the law professors (and Federalist Society members) William Baude and Michael Stokes Paulsen had put the argument into play. And I had read (not to mention heard, at length, on the phone) and took quite seriously what my friends Judge J. Michael Luttig and Professor Laurence H. Tribe had to say about it here in The Atlantic—that the Fourteenth Amendment clearly commands, in plain language, that Trump never hold federal office again.

[David Frum: The Colorado Supreme Court just gave Republicans a chance to save themselves]

Their points were strong. But much as I never want to see Trump near the White House again, I wasn’t quite buying them. The argument seemed somehow too good to be true. And frankly, from a political standpoint, it would be better for the country if Trump were thrashed at the polls, as I think he ultimately would be. There had to be a wrinkle. I just knew it.

But last night changed my mind. Not because of anything the Colorado Supreme Court majority said. The three dissents were what convinced me the majority was right.

The dissents were gobsmacking—for their weakness. They did not want for legal craftsmanship, but they did lack any semblance of a convincing argument.

For starters, none of the dissents challenged the district court’s factual finding that Trump had engaged in an insurrection. None of the dissents seriously questioned that, under Section 3 of the Fourteenth Amendment, Trump is barred from office if he did so. Nor could they. The constitutional language is plain. You can’t be president if you previously took an oath “as an officer of the United States … to support the Constitution of the United States” but “engaged in insurrection or rebellion” against, or have “given aid or comfort to the enemies of,” that Constitution or the nation it charters.

Nor did the dissents challenge the evidence—adduced during a five-day bench trial, and which, three years ago, we saw for ourselves in real time—that Trump had engaged in an insurrection by any reasonable understanding of the term. And the dissenters didn’t even bother with the district court’s bizarre position that even though Trump is an insurrectionist, Section 3 doesn’t apply to him because the person holding what the Constitution itself calls the “Office of the President” is, somehow, not an “officer of the United States.”

[Adam Serwer: The Colorado Ruling Calls the Originalists’ Bluff]

Instead, the three dissenters mostly confined themselves to saying that state law doesn’t provide the plaintiffs with a remedy. But that won’t help Trump. This case seems headed for the Supreme Court of the United States, which has no authority to make definitive pronouncements about state law. In Colorado, the Supreme Court of Colorado has the last word on that. And it now has spoken.

Yet even the dissenters’ contentions about state law made little sense. Chief Justice Brian Boatright argued that, while Colorado law requires its secretary of state to examine the constitutional qualifications of presidential candidates, it doesn’t allow her to consider whether they are constitutionally disqualified.

Nothing in the state statute suggests that’s the case, and it’s plainly illogical. Every qualification necessarily establishes a disqualification. If the Constitution says, as it does, that you have to be 35 years of age to serve as president, you’re out of luck—disqualified—if you’re 34 and a half. By the same token, if you’ve engaged in an insurrection against that Constitution in violation of your oath to it, you’ve failed to meet the ironclad (and rather undemanding) requirement that you not have done that.

Boatright’s suggestion that the insurrection issue presents something too complex for Colorado’s election-dispute-resolution procedures is equally unconvincing. Reviewing the tabulation of statewide votes can be complicated—remember those Florida “chads” in 2000?—but the courts have to get it done, and quickly. It’s hard to imagine that assessing the undisputed record of Trump’s miscreance presents any more complexity than that.

And no stronger is Justice Carlos Samour’s suggestion that Trump was somehow deprived of due process by the proceedings in the district court. This was a full-blown, five-day trial, with sworn witnesses and lots of documentary exhibits, all admitted under the traditional rules of evidence before a judicial officer, who then made extensive written findings of fact under a stringent standard of proof. Every day in this country, people go to prison—for years—with a lot less process than Trump got here. As for the expeditiousness of the proceedings, that’s in the very nature of election disputes: Recall, once again, Florida in 2000. And Samour’s suggestion that Trump was denied a fair trial because he didn’t have a jury is almost embarrassing: Any first-year law student who has taken civil procedure could tell you that election cases are not even close to the sort of litigation to which a Seventh Amendment jury-trial right would attach.

[Donald Ayer: The Trump Prosecutions Are Cause to Celebrate the Rule of Law]

The closest the dissents come to presenting a federal-law issue that ought to give someone pause comes in Samour’s argument that Section 3 is not self-executing—that it can’t be enforced unless Congress passes a law detailing how it can be enforced. The majority opinion, though, along with Paulsen and Baude and Luttig and Tribe, have disposed of that argument many times over. All you need to do is to look, as any good Scalia-like textualist would, to the words and structure of the Fourteenth Amendment.

True, Section 5 of the amendment gives Congress the power to enact enforcement legislation. But nothing in the amendment suggests that such legislation is required—that Section 3 (or any other prohibition in the amendment) has no teeth unless Congress implants them. To hold otherwise would mean that Section 1 of the Fourteenth Amendment—which contains the more familiar prohibitions against state deprivations of equal protection and due process—would likewise have been born toothless. Which would mean that, if every federal civil-rights statute were repealed tomorrow, states could immediately start racially resegregating their schools. That’s not the law, and thankfully so.

So the dissents showed one thing clearly: The Colorado majority was right. I dare not predict what will happen next. But if Trump’s lawyers or any members of the United States Supreme Court want to overturn the decision, they’d better come up with something much, much stronger. And fast.