concurring.
I join the Court’s opinion but write separately to set forth my understanding that the Court does not hold that an appellate court can fulfill its obligations of meaningful review by simply reciting the formula for harmless error. In Chapman v. California, 386 U. S. 18 (1967), we held that before a federal constitutional error can be held harmless, the reviewing court must find “beyond a reasonable doubt that the error complained of did not contribute to the verdict obtained.” Id., at 24. This is a justifiably high standard, and while it can be met without uttering the magic words “harmless error,” see ante, at 540, the reverse is not true. An appellate court’s bald assertion that an error of constitutional dimensions was “harmless” cannot substitute for a principled explanation of how the court reached that conclusion. In Clemons v. Mississippi, 494 U. S. 738 (1990), for example, we did not hesitate to remand a case for “a detailed explanation based on the record” when the lower court failed to undertake an explicit analysis supporting its “cryptic,” one-sentence conclusion of harmless error. Id., at 753. I agree with the Court that the Florida Supreme Court’s discussion of the proportionality of petitioner’s sentence is not an acceptable substitute for harmless error analysis, see ante, at 539-540, and I do not understand the Court to say that the mere addition of the words “harmless error” would have sufficed to satisfy the dictates of Clemons.