Court rules 5-4 against nullification by state courts
By 
John Cruz was convicted of murder in Arizona, after a judge informed a jury that if they didn’t return a death penalty sentence that he would have to decide between life without parole or life with parole. This was a lie — Arizona law did not permit parole for 1st degree murder. And therefore the death sentence voted by the jury violated binding Supreme Court precedent which requires that juries be informed that the alternative to a death sentence is life without parole if the state has cited the possibility of recidivism as a justification for the death penalty. The Arizona courts, however, simply refused to adhere to the Court’s holding, claiming that there was a specious “independent state justification” for maintaining the sentence.
By a narrow majority, with Roberts and Kavanaugh joining the three liberals, the Court did not permit this particular Catch-22:
While the particular case of Simmons arose out of South Carolina, the rule that it announced—states must inform juries whether a capital defendant will be eligible for release if they are not sentenced to death—applies everywhere. But Arizona courts refused to apply the rule, even after Arizona changed its law so that people who are convicted of capital crimes were not eligible for parole. Arizona courts refused to apply Simmons for almost two decades, until the court shot them down. In 2016, in Lynch v. Arizona, the Supreme Court held that Arizona was bound by the Supreme Court’s decisions—and that Arizona therefore had to inform Arizona juries whether capital defendants would be eligible for parole.
But the Arizona courts didn’t give up. Instead, Arizona courts refused to apply the Supreme Court’s decision in Lynch to cases that had already been decided. They refused to apply Lynch on the grounds that state law allowed defendants to challenge their convictions or sentences on the basis of “new” Supreme Court rules. And, Arizona continued, Lynch did not announce a “new” rule. Lynch had simply applied an existing rule (from Simmons) to Arizona without actually changing the law in the process.
What? Arizona’s position would have effectively left Arizona defendants with no remedy at all. That’s because before Lynch, Arizona courts were refusing to apply the rule in Simmons. And after Lynch, Arizona courts maintained that they still didn’t have to apply Simmons to those older cases, because it was so obvious that they should have been applying Simmons all along.
As Justice Elena Kagan explained at oral argument:
I think Kafka would have loved this. Cruz loses his Simmons claims on direct appeal because the Arizona courts say point-blank Simmons has never applied in Arizona. And then he loses the next time around because the Arizona courts say Simmons always applied … I mean, tails you win, heads I lose, whatever that expression is? I mean, how—how can you run a railroad that way?
Indeed. In a majority opinion by Justice Sonia Sotomayor, the Supreme Court affirmed that states cannot play “heads I win, tails you lose” with constitutional rights. (Sotomayor wrote for a majority of five that included the other Democratic appointees, Chief Justice John Roberts, and Justice Brett Kavanaugh. Justice Amy Coney Barrett wrote the dissent. This is basically the same lineup, with Justice Ketanji Brown Jackson replacing Justice Stephen Breyer, as the case last term where the court preserved the procedure that allowed people sentenced to death to challenge how the state planned to execute them.)
Had the Supreme Court countenanced Arizona’s scheme, it would have enabled states to ignore Supreme Court cases that Arizona didn’t like. It would have permitted states to refuse to give effect to any Supreme Court precedent the states and Supreme Court justices didn’t like—and to deny people their rights in the process.
It’s good the case came out correctly, but he fact that Arizona’s position got one vote, let alone four, is outrageous. Kim Scheppele’s article about the role judicially manufactured Catch-22s played in Bush v. Gore remains highly relevant.