The Interaction Between Victim Race and Gender on Sentencing Outcomes in Capital Murder Trials

• Date: 01 May 2006
• DOI: 10.1177/1088767905285510
• Author: Sondra J. Fogel,Beth Bjerregaard,Amy R. Stauffer,John K. Cochran,M. Dwayne Smith
• Published date: 01 May 2006
• Subject Matter: Articles

10.1177/1088767905285510
Homicide Studies
Stauffer et al. / Race and Gender in Sentencing
Homicide Studies
Volume 10 Number 2
May 2006 98-117
© 2006 Sage Publications
The Interaction Between Victim
10.1177/1088767905285510
http://hs.sagepub.com
Race and Gender on Sentencing
hosted at
http://online.sagepub.com
Outcomes in Capital Murder
Trials
A Further Exploration
Amy R. Stauffer
University of Florida, Gainesville
M. Dwayne Smith
John K. Cochran
Sondra J. Fogel
University of South Florida
Beth Bjerregaard
University of North Carolina–Charlotte
This study extends previous research on the interactive effects of victim race and gender
on death sentence outcomes reported by Williams and Holcomb (2004). They report an
interactive effect between victim race and victim gender on Ohio death sentencing out-
comes, such that killers of White women are especially at risk of receiving death sen-
tences. The study here seeks to determine if the Williams and Holcomb finding holds for
a sample of murder cases in North Carolina for which the state sought the death penalty.
Initial results of a descriptive analysis suggest a White female victim effect, but the intro-
duction of control variables via logistic regression equations yields no gender or race
interactions as predictors of sentencing outcomes. Reasons for the different outcomes are
explored, and topics requiring further exploration are discussed.
Keywords: capital punishment; race; gender
Oneofthemostpersistentfindingsofstudiesconcerningpredictorsofcontempo-
rary death sentencing in the United States has been a so-called “White victim”
effect. In essence, research has shown the killers of Whites are at escalated risk of
receiving death sentences, even when controlling for a host of other legal and extrale-
gal variables that might influence the outcome. For an extensive review of this litera-
ture and a discussion of possible explanations for the effect, readers are referred to
Baldus and Woodworth (2003).
98

---

Stauffer et al. / Race and Gender in Sentencing
99
In a 2004 article in Homicide Studies, Williams and Holcomb argue that the White
victim effect may be more complex than has been recognized because previous litera-
ture has typically failed to take into account the possible interactive effects of race and
gender. That is, although the murders of White victims are thought to evoke more pub-
lic outrage and thus a heightened likelihood of a death sentence, a similar effect may
exist for the murders of women, especially under conditions likely to generate a
charge of capital murder. The results they present support this contention, showing
that an apparent White victim effect in death sentencing found for murders in Ohio is
actually a “White female victim” effect. In essence, the murders of White female vic-
tims are more likely to result in death sentences as compared to other race or gender
combinations. In contrast, cases with male victims do not generate a racial disparity in
sentencing outcomes.1
The purpose of the research reported here is to determine whether Williams and
Holcomb’s (2004) finding can be reproduced with data from North Carolina when
restricted to a sample of cases in which the state actually sought the death penalty. It is
important to note that because of differences in the data discussed below, we do not
purport to replicate the Williams and Holcomb study. Instead, we seek to determine
whether the dynamics discussed in their article similarly influence death sentencing in
North Carolina.
The Williams and Holcomb (2004) Study
It is prudent to further discuss the Williams and Holcomb (2004) study before pro-
ceeding to a description of the present research. The essence of their argument is that
race has been a special focus of capital sentencing outcomes in the era following
Gregg v. Georgia (1976). Although evidence of a persistent race-of-offender effect
across jurisdictions has been mixed, a more consistent finding has been that killers of
White victims are more likely to receive a death sentence than the killers of victims
from other races, even when controlling for legal, crime-specific, and other extralegal
variables. One of the best known of these studies was that of Baldus, Woodworth, and
Pulaski (1990) in which a (White) race-of-victim effect was shown to hold for a Geor-
gia death sentencing, even when controlling for more than 200 other variables con-
ceivably affecting such outcomes. However, Williams and Holcomb rightly note that a
less discussed finding in much of this literature is a gender-of-victim effect that tilts
death sentencing toward the murderers of female victims (for recent research that fur-
ther supports this contention, see Curry, Lee, & Rodriguez, 2004). The practice has
been to consider these as independent influences on sentencing outcomes, a practice
that Williams and Holcomb believe may obscure interactive effects between these two
variables. In particular, they believe that historical and cultural forces operate to view
the murder of a White female being particularly egregious and deserving of severe
Authors’ Note: This is a revised and expanded version of an article presented by the first author at the 2004
American Society of Criminology meetings. Please address correspondence to M. Dwayne Smith, Depart-
ment of Criminology, University of South Florida, 4202 E. Fowler Ave., SOC107, Tampa, FL 33620; e-mail:
mdsmith@cas.usf.edu.

---

100
Homicide Studies
sanction.2 Similar forces contribute to a devaluation of Black males, resulting in their
murders being perceived as the least threatening to the larger society and therefore less
likely to warrant a severe sanction such as the death penalty. Thus, Williams and
Holcomb believe that the apparent racial and gender influences in capital sentencing
may be best understood in terms of their potentially interactive effects and may
involve more complex interpretations than their independent effects would suggest.
To test their argument, Williams and Holcomb (2004) analyze cases of murder in
Ohio as reported in the Federal Bureau of Investigation’s Supplementary Homicide
Reports (SHR) for the years 1981 through 1994.3 Excluding missing data, 5,320 cases
were available for analysis, 271 of which were determined to result in a death sentence
for the offender. Stressing that their study did not address the issue of case selection for
capital prosecution, Williams and Holcomb sought to determine factors that distin-
guished the death sentence cases from those with other outcomes. Using logistic
regression, 14 control variables, including 3 victim and race interactive terms (White
male, Black male, Black female, with White female as the reference category) were
used in several combinations. Across different equations with varying factors, a statis-
tically significant effect was shown for cases having a White female victim, with a
death sentence more likely to result. Although this effect was weak compared to other
factors (e.g., the case involved multiple victims), it persisted across different sets of
controls.
In a concluding comment, Williams and Holcomb (2004) wrote, “call on interested
scholars with access to more complete data to test our hypotheses and challenge our
findings” (p. 372). Intrigued by both the premise and results of their research, we do so
here.
The Present Study
At the outset, we stress that the present study is not intended as a replication of
Williams and Holcomb (2004). As detailed below, although pursuing the same
research questions, there are two important differences in our studies.
First, the data for our study are from North Carolina rather than from Ohio. North
Carolina is a particularly relevant state for death penalty research because it is an
active state in terms of pursuing capital punishment; as of late 2005, North Carolina
ranked 6th in the number of individuals awaiting execution, 7th in the number of exe-
cutions carried out post-Gregg, 10th in the ratio of murders to death sentences, and
13th in state execution rates per 10,000 population (Death Penalty Information Center,
2005). Because it is a southern and former Confederate state where slavery was an
established institution, it could be expected that the historical racial animus thought to
condition perceptions of crime by and against people of color would be even more pro-
nounced than in a midwestern state such as Ohio (see Corzine, Huff-Corzine, & Whitt,
1999). As well, historical traditions of chivalry that proscribe offenses against
women—particularly White women—would be assumed to elevate sanctions against
their offenders. Therefore, it could be speculated that an interactive race or gender-of-

---

Stauffer et al. / Race and Gender in Sentencing
101
victim effect in death sentencing could be even more pronounced in North Carolina
than in Ohio.
A second difference is that although the Williams and Holcomb (2004) study traces
eventual death sentences from a sample of all Ohio homicides during a specified time
period, our study restricts analyses to a large sample of capital cases in North Carolina
from the post-Gregg era. By capital cases, we mean those in which (a) a first-degree
murder conviction was secured, (b) the state sought the death penalty, and (c) the trial
advanced to a sentencing phase whereby the jury recommended either a life sentence
or the death penalty for the defendant.
In adopting this focus, we recognize that restricting our analyses to capital trials
excludes other segments of the criminal justice process in which racial and/or gender
biases could...