Cite as: 555 U. S. ____ (2008) 
             1

                    Statement of STEVENS, J. 


SUPREME COURT OF THE UNITED STATES
                DOUGLAS OLIVER KELLY
07–11073                 v.
                    CALIFORNIA

                    SAMUEL ZAMUDIO
07–11425                   v.
                      CALIFORNIA
 ON PETITIONS FOR WRITS OF CERTIORARI TO THE SUPREME
                 COURT OF CALIFORNIA
     Nos. 07–11073 and 07–11425.   Decided November 10, 2008


   The petitions for writs of certiorari are denied. JUSTICE
SOUTER would grant the petition for a writ of certiorari in
No. 07–11073.
   Statement of JUSTICE STEVENS respecting the denial of
the petitions for writs of certiorari.
   These two capital cases raise questions concerning the
admissibility of so-called “victim impact evidence” during
the penalty phase of a capital trial. The term is a misno
mer in capital cases because the evidence does not de
scribe the impact of the crime on the victim—his or her
death is always an element of the offense itself. Rather, it
describes the impact of the victim’s death on third parties,
usually members of the victim’s family.
   In the first of these cases, petitioner Douglas Kelly was
convicted of murdering 19-year-old Sara Weir. 

, 

(2007). The prosecution played a 20
minute video consisting of a montage of still photographs
and video footage documenting Weir’s life from her infancy
until shortly before she was killed. The video was nar
rated by the victim’s mother with soft music playing in the
background, and it showed scenes of her swimming,
horseback riding, and attending school and social func
tions with her family and friends. The video ended with a
view of her grave marker and footage of people riding
2                      KELLY v. CALIFORNIA

                       Statement of STEVENS, J.

horseback in Alberta, Canada—the “ ‘kind of heaven’ ” in
which her mother said she belonged. See 

171 P. 3d, at 557–558.1
    In the second case, petitioner Samuel Zamudio was
convicted of robbing and murdering Elmer and Gladys
Benson. 

, 

(2008). Two of the
victims’ daughters and two of their grandchildren testified
about the effects of the murders on themselves and their
families. During one daughter’s testimony the prosecution
played a video containing 118 photographs of the victims
at various stages of their lives, including their childhood
and early years of marriage. The photographs showed the
couple raising their children, serving in the military,
hunting, fishing, vacationing, bowling, celebrating holi
days and family events, and attending recognition dinners
for Gladys’s community service. “The last three photo
graphs in the montage showed, in order, Gladys’ grave
marker with the inscription readable, Elmer’s grave
marker with the inscription readable, and both grave
markers from a distance, each accompanied by a vase of
flowers.” 

181 P. 3d, at 134.
   In both cases the California Supreme Court upheld the
admissibility of the videos. The court explained that the
video admitted during Kelly’s sentencing “expressed no
outrage” and contained no “clarion call for vengeance,” but
“just implied 

, 171 P. 3d, at
558. Similarly, the court held that the video shown during
Zamudio’s penalty phase proceedings was “ ‘not unduly
emotional.’ 

, 181 P. 3d, at 137. Only
one dissenting justice expressed any concern that the
evidence had the potential to “imbue the proceedings with
‘a legally impermissible level of emotion.’ 

, 171 P. 3d, at 575 (Moreno, J., concurring and dissent
——————
  1 The full video is available online at http://www.supremecourtus.gov/

opinions/video/kelly_v_california.html and in Clerk of Court’s case file.
                     Cite as: 555 U. S. ____ (2008)                    3

                       Statement of STEVENS, J.

ing). No member of the court suggested that the evidence
shed any light on the character of the offense, the charac
ter of the offender, or the defendant’s moral culpability.

                                   I
   Victim impact evidence made its first appearance in this
Court’s jurisprudence in 1987. Booth v. Maryland, 

(1987).2 In earlier landmark cases, such as
Williams v. New York, 

(1949), and Lockett v.
Ohio, 

(1978), evidence probative of the cul
pability and character of the offender and the circum
stances of the offense had marked the outer limits of the
kind of evidence admissible in capital sentencing. Consis
tent with that precedent, in our first encounter with vic
tim impact evidence, the Court announced a rule that
categorically “prohibit[ed] a capital jury from considering
victim impact evidence” that “described the personal
characteristics of the victims and the emotional impact of
the crimes on the family.” 

–502. It
was the unique character of the death penalty that justi
fied Booth’s per se rule: The opinion relied on the fact that
death is a “punishment different from all other sanctions,”

——————
  2 Victim impact evidence is a category unmentioned by Wigmore’s

treatise or other classic works on the law of evidence. Its inclusion in
both capital and noncapital cases is a phenomenon of recent origin,
arising out of the victims’ rights movement of the late 1970’s. See
Carrington & Nicholson, The Victims’ Movement: An Idea Whose Time
Has Come, 11 Pepperdine L. Rev. 1, 8 (Symposium 1984) (describing
early victories of the victims’ rights movement, including passage of the
Omnibus Victim and Witness Protection Act of 1982, 96 Stat. 1248,
which mandated the inclusion of victim impact statements in federal
presentence reports); MacDonald, Towards a Bicentennial Revolution
in Criminal Justice: The Return of the Victim, 13 Am. Crim. L. Rev.
649, 670 (1975—1976) (describing early “innovative” attempts to
integrate victims into the sentencing process).
4                  KELLY v. CALIFORNIA

                   Statement of STEVENS, J.

n. 12, and on our earlier admonition that any
decision to impose the death sentence must “ ‘be, and
appear to be, based on reason rather than caprice or emo
tion,’ ” 

(quoting Gardner v. Florida, 

, 358 (1977) (opinion of STEVENS, J.)).
   Throughout the late 1970’s and for much of the follow
ing decade, the fact that “death is a different kind of pun
ishment from any other that may be imposed in this coun
try,” 

had justified placing limits on its
permissible applications, see, e.g., Godfrey v. Georgia, 

, 433 (1980) (plurality opinion), and requiring
special procedural protections for the defendant, see

(plurality opinion). Our decision
in Booth flowed naturally from the same principle.
   Beginning in the late 1980’s, however, changes in the
Court’s capital jurisprudence began to weaken the proce
dural and substantive safeguards on which we had earlier
insisted. In Tison v. Arizona, 

(1987), rather
than adhere to the rule announced in Enmund v. Florida,

(1982), which prohibited death sentences for
defendants who neither killed nor intended to kill a vic
tim, a majority of the Court held that felony murder could
qualify as a capital offense. Soon thereafter, the Court
rejected a challenge to a death sentence based on evidence
that a victim’s race enhanced the likelihood that a Georgia
jury would impose the death penalty. McCleskey v. Kemp,

(1987). As Justice Blackmun presciently
observed, the fact that “death is different” was fast becom
ing a justification for applying “a lesser standard of scru
tiny” in capital cases. See 

348 (dissenting
opinion).
   Confirming that observation, the Court’s 1991 opinion in
Payne v. Tennessee, 

, overruled Booth in
short order, giving prosecutors a powerful new weapon in
capital cases. At issue in Payne was the admission of
penalty phase testimony by the mother of a deceased
                 Cite as: 555 U. S. ____ (2008)           5

                   Statement of STEVENS, J.

victim. The woman testified about the effect of the crime
on her surviving grandson, who had witnessed the murder
of his mother and baby sister and had himself nearly been
killed by the same attack. Her testimony powerfully
conveyed her grandson’s suffering, but “she[d] no light on
the defendant’s guilt or moral culpability.” 

(STEVENS, J., dissenting). By its very poignancy, the
testimony “encourage[d] jurors to decide in favor of death
rather than life on the basis of their emotions rather than
their reason.” 

inherent danger
posed by such testimony, the Court rejected Booth’s per se
rule barring the admissibility of victim impact evidence in
capital proceedings. Declaring such evidence to be “simply
another form or method of informing the sentencing au
thority about the specific harm caused by the crime in

, the Court held that prosecu
tors should be permitted to present evidence “offering a
quick glimpse of the life which [the] defendant chose to
extinguish” and “demonstrating the loss to the victim’s
family and to society . . . result[ing] from the defendant’s
homicide,” 

(internal quotation marks omitted).
   Given Payne’s sharp retreat from prior precedent, it is
surprising that neither the opinion of the Court nor any of
the concurring opinions made a serious attempt to define
or otherwise constrain the category of admissible victim
impact evidence. Instead, the Court merely gestured
toward a standard, noting that, “[i]n the event that evi
dence is introduced that is so unduly prejudicial that it
renders the trial fundamentally unfair, the Due Process
Clause of the Fourteenth Amendment provides a mecha
nism for relief.” 

That statement represents
the beginning and end of the guidance we have given to
lower courts considering the admissibility of victim impact
evidence in the first instance.
6                      KELLY v. CALIFORNIA

                       Statement of STEVENS, J.

                                  II
  In the years since Payne was decided, this Court has left
state and federal courts unguided in their efforts to police
the hazy boundaries between permissible victim impact
evidence and its impermissible, “unduly prejudicial”
forms. Following Payne’s model, lower courts throughout
the country have largely failed to place clear limits on the
scope, quantity, or kind of victim impact evidence capital
juries are permitted to consider. See generally, Logan,
Through the Past Darkly: A Survey of the Uses and
Abuses of Victim Impact Evidence in Capital Trials, 

. Rev. 143 (1999). Not only have courts allowed
capital sentencing juries to hear brief oral or written
testimony from close family members regarding victims
and the direct impact of their deaths; they have also al
lowed testimony from friends, neighbors, and co-workers
in the form of poems, photographs, hand-crafted items,
and—as occurred in these cases—multimedia video pres
entations. See Blume, Ten Years of Payne: Victim Impact
Evidence in Capital Cases, 88 Cornell L. Rev. 257, 271–
272 (2003) (collecting cases).
  Victim impact evidence is powerful in any form.3 But in
each of these cases, the evidence was especially prejudi
cial. Although the video shown to each jury was emotion
——————
  3 As one Federal District Judge put it, “I cannot help but wonder if

Payne . . . would have been decided in the same way if the Supreme
Court Justices in the majority had ever sat as trial judges in a federal
death penalty case and had observed first hand, rather than through
review of a cold record, the unsurpassed emotional power of victim
impact testimony on a jury. It has now been over four months since I
heard this testimony [in a codefendant’s case] and the juror’s sobbing
during the victim impact testimony still rings in my ears. This is true
even though the federal prosecutors in [the case] used admirable
restraint in terms of the scope, amount, and length of victim impact
testimony.” United States v. Johnson, 

, 1107 (ND
Iowa 2005).
                 Cite as: 555 U. S. ____ (2008)            7

                   Statement of STEVENS, J.

ally evocative, it was not probative of the culpability or
character of the offender or the circumstances of the of
fense. Nor was the evidence particularly probative of the
impact of the crimes on the victims’ family members: The
pictures and video footage shown to the juries portrayed
events that occurred long before the respective crimes
were committed and that bore no direct relation to the
effect of crime on the victims’ family members.
   Equally troubling is the form in which the evidence was
presented. As these cases demonstrate, when victim
impact evidence is enhanced with music, photographs, or
video footage, the risk of unfair prejudice quickly becomes
overwhelming. While the video tributes at issue in these
cases contained moving portrayals of the lives of the vic
tims, their primary, if not sole, effect was to rouse jurors’
sympathy for the victims and increase jurors’ antipathy
for the capital defendants. The videos added nothing
relevant to the jury’s deliberations and invited a verdict
based on sentiment, rather than reasoned judgment.
   I remain convinced that the views expressed in my
dissent in Payne are sound, and that the per se rule an
nounced in Booth is both wiser and more faithful to the
rule of law than the untethered jurisprudence that has
emerged over the past two decades. Yet even under the
rule announced in Payne, the prosecution’s ability to admit
such powerful and prejudicial evidence is not boundless.
   These videos are a far cry from the written victim im
pact evidence at issue in Booth and the brief oral testi
mony condoned in Payne. In their form, length, and scope,
they vastly exceed the “quick glimpse” the Court’s major
ity contemplated when it overruled Booth in 1991. At the
very least, the petitions now before us invite the Court to
apply the standard announced in Payne, and to provide
the lower courts with long-overdue guidance on the scope
of admissible victim impact evidence. Having decided to
tolerate the introduction of evidence that puts a heavy
8                  KELLY v. CALIFORNIA

                   Statement of STEVENS, J.

thumb on the prosecutor’s side of the scale in death cases,
the Court has a duty to consider what reasonable limits
should be placed on its use.