It's an ugly case. You can see why the jury sentenced Mr. Petrocelli to death:
"On March 29, 1982, Petrocelli went on a test drive of a
Volkswagen pickup truck with James Wilson, a used car
salesman, in Reno, Nevada. At some point during that test
drive, Petrocelli shot and killed Wilson. . . .
Nearly a year before killing Wilson, in May 1981,
Petrocelli had pleaded guilty in Washington State to
kidnaping his girlfriend, Melanie Barker. He had received a
suspended sentence conditioned on his completion of a drug
treatment program. Petrocelli absconded from the treatment
program twice and never completed it. Petrocelli shot and
killed Barker in Washington State in October 1981, five
months before he killed Wilson in Nevada."
Bad facts for the defendant, for sure.
But the panel unanimously reverses the death sentence due to the admission of an expert's testimony at the penalty phase. And this is not an isolated occurrence. As the penultimate paragraph of Judge Fletcher's opinion recounts:
"We have encountered Dr. Gerow before. He testified for
the prosecution in Sechrest in very much the same manner he
testified for the prosecution in the case before us. Gerow
testified that Sechrest “was an incurable sociopath” who was
“extremely dangerous and could not be rehabilitated.”
Sechrest, 549 F.3d at 813. We held in Sechrest that the
combined effect of Gerow’s testimony and an instruction
identical to Instruction 5 “had a substantial influence on the
jury’s decision to sentence Sechrest to death.” Id. We
similarly conclude, in this case, that Gerow’s improperly
admitted testimony, understood in the light of Jury Instruction
5, “had [a] substantial and injurious effect or influence in
determining the jury’s verdict.”"
Judge Christen's position is perhaps even stronger than the majority. She says in her concurrence:
"I agree that Petrocelli’s death sentence must be reversed.
I write separately because, in my view, even if the State could
show that the prosecutor’s tactics had not prejudiced the
jury’s verdict, Petrocelli’s case is one of the very few in
which deliberate prosecutorial misconduct and egregious trial
errors warrant habeas relief. See Brecht v. Abrahamson,
507 U.S. 619, 638 n.9 (1993) (stating that a deliberate and
especially egregious trial error, or one that is combined with
a pattern of prosecutorial misconduct, might warrant habeas
relief, even if the jury’s verdict is not substantially
influenced). Brecht’s footnote nine is rarely employed, but
the Fifth and Seventh Circuits have each relied on it one time
in cases where an error (or errors) did not easily fit into either
the “structural error” or “trial error” category. The errors in
Petrocelli’s case were equally pervasive, flouted Supreme
Court authority, and undermined the integrity of the criminal
justice process."
Strong stuff.