Cite as: 527 U. S. 373 (1999)
Ginsburg, J., dissenting
Appellate courts should hesitate to assert confidence that "elimination of improperly considered aggravating circumstances could not possibly affect the balance." Barclay v. Florida, 463 U. S. 939, 958 (1983). Adding the overlapping aggravators to the more disturbing misinformation conveyed in the charge, I see no basis for concluding " 'it would have made no difference if the thumb had been removed from death's side of the scale.' " 132 F. 3d, at 251 (quoting Stringer v. Black, 503 U. S. 222, 232 (1992)).
V
The Fifth Circuit's tolerance of error in this case, and this Court's refusal to face up to it, cannot be reconciled with the recognition in Woodson v. North Carolina, 428 U. S. 280, 305 (1976) (plurality opinion), that "death is qualitatively different." If the jury's weighing process is infected by the trial court's misperceptions of the law, the legitimacy of an ensuin opposition, the Government did not challenge the Fifth Circuit's determination of error, but focused solely on whether the error was harmless. Justice Thomas, writing for a plurality, nevertheless addresses the Government's newly raised argument. See ante, at 395-402. I would hold the Government to a tighter rein and dismiss the tardy argument as waived. See Roberts v. Galen of Va., Inc., 525 U. S. 249, 253 (1999) (per curiam); South Central Bell Telephone Co. v. Alabama, 526 U. S. 160 (1999); cf. this Court's Rule 15.2.
It is evident that the issue held back by the Government was not "predicate to an intelligent resolution of the question presented." Ohio v. Robinette, 519 U. S. 33, 38 (1996) (internal quotation marks omitted). But see ante, at 397, n. 12. Justice Thomas treats the two issues as separate and independent. He maintains first that there was no error. Writing for the Court, he then proceeds to assume there was error and concludes that any error was harmless. Either holding would do to support the Court's disposition. See, e. g., United States v. Hasting, 461 U. S. 499, 506, n. 4, 510-512 (1983) (holding presumed error harmless rather than deciding whether there was, in fact, error; Court explains "[t]he question on which review was granted assumed that there was error and the question to be resolved was whether harmless-error analysis should have applied"); id., at 512-513 (Stevens, J., concurring) (Court should decide case on the ground that there was no error, without reaching harmless-error question).
421
Page: Index Previous 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 NextLast modified: October 4, 2007