Abstract
Research exploring attitudes toward the death penalty is common in the field of criminal justice. Additionally, a substantial body of literature has examined public perceptions of sex offenders and punishment in the U.S. Unfortunately, few studies have sought to examine perceptions of the death penalty in relation to sexual offending. This study contributes to the literature by examining perceptions of the college students at a mid-sized university in the Southeastern United States as they relate to support for the death penalty in cases of sexual assault across victim age categories. Findings suggest that respondent perceptions are shaped by biological sex, political affiliation, college major, fear of crime, and parents’ level of education, and these relationships are uniform across victim age categories. Further, support for the death penalty appears inversely related with victim age.
Introduction
On March 2, 1998, 44-year-old Patrick O’Neal Kennedy called the police to report that his 8-year-old stepdaughter had been raped. According to reports, the extent of the victim’s injury was so severe that it required extensive emergency surgery. The injuries included a laceration to the left vaginal wall that resulted in a separation of the cervix from the back of the vagina, which in turn caused the rectum to protrude into the vaginal structure. It was also reported that “her entire perineum was torn from the posterior fourchette to the anus” (Kennedy v. Louisiana, 2008). Kennedy, himself, was ultimately arrested, charged and convicted of the offense. At the time of sentencing, the Louisiana state statute held that: “D. Whoever commits the crime of aggravated rape shall be punished by life imprisonment at hard labor without benefit of parole, probation, or suspension of sentence. “(1) However, if the victim was under the age of 12 years, as provided by Paragraph A(4) of this Section: “(a) And if the district attorney seeks a capital verdict, the offender shall be punished by death or life imprisonment at hard labor without benefit of parole, probation, or suspension of sentence, in accordance with the determination of the jury” (as cited in Kennedy v. Louisiana, 2008)
Based on this statute and given the severity of the victim’s injury, in 2003, a jury unanimously voted to sentence Kennedy to death. In upholding the decision of the lower court, the State Supreme Court argued that children were a protected class and sexual violence against them was unique in terms of the harm it inflicts. According to the Court, “[S]hort of first-degree murder, we can think of no other non-homicide crime more deserving [of capital punishment].” The case eventually made its way to the U.S. Supreme Court where, in the 2008 landmark decision in Kennedy v. Louisiana, the highest court in the land struck down the Louisiana state statute as unconstitutional. Although the Court acknowledged that occurrences of child rape are more frequent than intentional murder, it nonetheless ruled that the imposition of the death penalty for a rapist who did not kill the victim constitutes excessive punishment. In delivering the opinion of the 5-4 majority, Justice Kennedy noted that, “[b]ased both on consensus and our own independent judgment, our holding is that the death sentence for one who raped but did not kill a child, and who did not intend to assist another in killing the child, is unconstitutional under the Eight and Fourteenth Amendments” (Kennedy v. Louisiana, 2008).
To be clear, the U.S. Supreme Court decision in Coker v. Georgia (1977) had already prohibited the use of the death penalty for the rape of adult women. That Court declared that the use of the death penalty for rape was “grossly disproportionate and excessive.” Bell (2008) suggested that the justices’ interpretation of the Eighth Amendment in Coker was partially fashioned using the concept of “proportionality” established in Weems v. United States (1910). Delivering the opinion of the Court in Weems, Justice McKenna stipulated that the interpretation of the Eighth Amendment “will be regarded as a precept of justice that punishment for crime should be graduated and proportioned to the offense.” Based on precedent, the majority in Kennedy felt that despite the severity of the girl’s injury, the imposition of the death penalty for rape did not meet the proportionality test. Although the Kennedy ruling invalidated similar laws in Georgia, Montana, Oklahoma, South Carolina and Texas, the very existence of those capital child rape statutes suggests a sentiment among some that unlike adult rape, aggravated victimization of children is deserving of the most extreme punishment.
It is against this backdrop that the current study explores the factors that influence support for the death penalty in cases of rape and sexual assault. Specifically, using a sample of college students at a university in the southeastern U.S., the purpose of this study is to examine student perceptions of the death penalty in cases of sexual assault across victim age categories. Prior research suggests that the majority of U.S. residents may support application of the death penalty in sexual offenses involving children (Mancini & Mears, 2010; Quinnipiac University, 2008; Wiersma & Siedschlaw, 2016). These studies have not, however, explored the extent to which such support might vary between victim age categories. Given a newly formed 6-3 majority-conservative U.S. Supreme Court, coupled with long-standing public support for harsh punishment of sexual offenders, particularly those targeting children, (Burchfield et al., 2014; Meloy et al., 2013), such exploration is timely. Indeed, with the most conservative U.S. Supreme Court in nearly 70 years, states may seek to challenge the close decision in Kennedy. It is therefore important to further our understanding of how levels of support for application of the death penalty may vary by victim age.
Literature Review
Respondent Characteristics and Support for the Death Penalty
Research on the factors that influence support for the death penalty has long been a focal point in criminal justice research. Driven partially by the moral and ethical debates capital punishment generates, several theoretically relevant predictors of support for the death penalty have been explored. Much of the work on this topic has centered on understanding the influence of individual factors such as race/ethnicity. The general consensus among these studies is that Whites are more supportive of the death penalty than non-Whites (Barkan & Cohn, 1994; Bobo & Johnson, 2004; Cochran & Chamlin, 2006; Oliphant, 2018; Peffley & Hurwitz, 2007, 2010; Soss et al., 2003). A recent Pew Research finding revealed that more than half (59%) of Whites support the death penalty in cases of murder, compared to 47% of Hispanics and just 36% of Blacks (Oliphant, 2018).
In addition to race/ethnicity, sex has also emerged as a consistent and strong predictor of support for the death penalty, with males generally being more supportive of the death penalty in comparison to females (Applegate et al., 2002; Oliphant, 2018; Sandys & McGarrell, 1995; Skovron et al., 1989; Whitehead & Blankenship, 2000; Young, 1991). For example, Applegate et al. (2002) found that 81.9% of men supported the death penalty compared to 63.9% of women. A Pew Research Center poll also showed that men were significantly more likely to support the death penalty (61%) than oppose it (34%) (Oliphant, 2018). However, when it came to women, there was little difference between the percentage of those who supported capital punishment (46%) versus those who opposed it (45%) (Oliphant, 2018). An even more nuanced pattern emerges when race and gender intersect. Dotson and Carter (2012), for example, noted that while direct effects of gender reveal that 78.1% of men were in favor of the death penalty compared to 68.9% of women, White males (82.1%) and White females (73.2%) are more supportive of the death penalty than Black males (51.1%) and Black females (46.9%).
Likewise, the link between age and support for the death penalty is not clear. Recent public opinion polls show older adults as more likely than younger people to express support for the death penalty (Oliphant, 2018). Yet, other studies have found little evidence of a relationship (Cochran et al., 2003; Unnever & Cullen, 2006). However, using a more sophisticated statistical analysis, Anderson et al. (2017) detected a curvilinear relationship; compared to middle-aged adults, younger and older Americans were less likely to support the death penalty.
A similar trend emerges in terms of education. Fox et al. (1991), for example, found evidence of a potential non-linear relationship, with high school graduates being more likely to support the death penalty than those who did not complete high school, while those who are college educated are less likely to support the death penalty than high school graduates (Fox et al., 1991). Other studies suggest that lower levels of education tend to be associated with higher levels of support for the death penalty, whereas higher levels of education are typically associated with lower levels of support for the death penalty (Barkan & Cohn, 1994; Borg, 1997; Britt, 1998; Halim & Stiles, 2001; Vidmar & Ellsworth, 1974; Young, 1991). Specific to college students, Farnworth et al. (1998) and Lambert et al. (2008) found that criminal justice majors were more supportive of the death penalty compared to non-criminal justice majors. Other studies, however, reported no differences in support between criminal justice majors and non-criminal justice majors (Robbers, 2004; Schadt & DeLisi, 2007). Providing more insight into the understanding of the effects of education, Bohm et al. (1991) found that support for the death penalty diminishes the more one becomes educated on issues surrounding the death penalty.
Sociopolitical factors such as political affiliation have also been known to influence attitudes toward the death penalty. A majority of studies suggest political conservatism is more aligned with support for the death penalty in comparison to other political perspectives (Bobo & Johnson, 2004; Messner et al., 2006; Payne et al., 2004; Sandys & McGarrell, 1997; Unnever & Cullen, 2005, 2010; Wozniak & Lewis, 2010). The margins of support are relatively large with opinion poll numbers showing 77% of Republicans supporting capital punishment versus 35% of Democrats (Oliphant, 2018).
In addition, research finds that fear of crime is a significant predictor of punitiveness (Goodall et al., 2013; Hartnagel & Templeton, 2012; Johnson, 2009), including capital punishment (Costelloe et al., 2009; Unnever & Cullen, 2007; Wu et al., 2011). Some of the attention devoted to understanding the impact of fear of crime centers on punitiveness as an instrumental response to crime. This instrumental hypothesis holds that those who are afraid of crime see rising crime rate as a societal concern and view punishment as a deterrent are more likely to support harsh punishment such as the death penalty (Thomas & Foster, 1975; Tyler & Weber, 1982). Thomas and Foster (1975), for example, found a direct relationship between the perception of increasing crime rates and both fear of victimization and a willingness to employ harsh punishment as a response to criminal offenders. Contemporary studies have provided some support for this assertion. Anderson et al. (2017), for example, found evidence of a significant relationship between the violent crime rate and support for the death penalty.
In contrast, other studies have reported limited support for the instrumental perspective (Cohn et al., 1991; Hogan et al., 2005; Kleck & Jackson, 2017; Leverentz, 2011; Tyler & Weber, 1982; Welch et al., 2011). Additional works have found punitiveness to be associated with higher levels of media consumption rather than fear of crime (Kleck & Jackson, 2017; Rosenberger & Callanan, 2011). Furthermore, there is some evidence that this relationship may vary across race, as Hogan et al. (2005) and Cohn et al. (1991) found a significant relationship between fear of crime and support for punitiveness for White respondents, but no relationship among Black respondents. These inconsistencies prompted Tyler and Weber (1982) to posit that the lack of support for the instrumental perspective can, “best be explained as a direct result of beliefs in the value of the death penalty, beliefs which are closely associated with political-social attitudes” (p. 40).
Research examining the relationship between socio-economic status and punitive attitudes toward criminals has been inhibited by the complexity of the operationalization of the concept. Whether measured in terms of household income, employment status or highest level of educational attainment, the results of these studies are mixed. Some studies, for example, have found no evidence that perception of death penalty is influenced by the socio-economic status of respondents (Adinkrah & Clemens, 2018; Cochran & Chamlin, 2006; Johnson, 2001; see also Unnever & Cullen, 2007). Other studies suggest a more nuanced relationship. Costelloe et al. (2009), for example, found lower income white males are significantly more likely than their wealthier counterparts to embrace punitive attitudes toward criminals. Consistent with the economic insecurity hypothesis, they found elevated support for punitiveness among respondents who anticipated a deterioration of their financial situation in the very near future.
In addition to respondent characteristics, several studies suggest that support for the death penalty in cases of murder may be further influenced by aggravating factors surrounding the event (Durham et al., 1996). With respect to victim age, for example, Marier et al. (2018) suggest that victim age is an indicator of perceived victim vulnerability, particularly among child victims, and is a strong predictor of whether or not jurors recommend the death penalty in cases of murder. Additional studies point to higher levels of support for the death penalty in cases involving female and White victims (Holcomb et al., 2004), particularly those involving Black male offenders (Cochran et al., 2019; Paternoster, 1984; Radelet & Pierce, 1991), as well as instances in which the murder involved rape or sexual assault (Jennings et al., 2017; Richards et al., 2016).
Sex Offenders, Punishment, and the Death Penalty
The 1990s were labeled as “the decade of the predatory sex offender” (Nash, 1999, p. 45). Despite a decline in reported sex offenses, the period ushered in a push for harsher punishments for sex offenders in response to several highly publicized cases (Calkins et al., 2014; Cohen & Jeglic, 2007; Levenson & Cotter, 2005; Petrunik, 2002). Many of these involved child victims and murder, often by a perpetrator with prior sex crime convictions (Vasquez et al., 2008). Spurred by public support, more punitive sex offender laws were legislated (Jenkins, 1998; Meloy et al., 2013; Vess, 2009), including the use of registries, community notification, residency restrictions, and electronic monitoring. Further, despite the abject failure of these initiatives to reduce the frequency of sexual victimization (e.g., Dierenfeldt & Carson, 2017; Vasquez et al., 2008), they remain popular with the public as a consequence of a perpetual panic that surrounds sex offenders (Burchfield et al., 2014). Indeed, multiple studies illustrate the fear and outrage associated with sex offenders (Brown et al., 2008; Craun & Theriot, 2009; Kernsmith et al., 2009; Levenson et al., 2007; Olver & Barlow, 2010; Schiavone & Jeglic, 2009), a phenomenon that likely stems, in part, from misperceptions concerning their offending patterns and recidivism rates (Calleja, 2015; Cortoni et al., 2010; Hanson & Bussiere, 1998; Hanson & Harris, 2001; Hanson et al., 2003; Meloy et al., 2008; Sample & Bray, 2003; Vess & Skelton, 2010; Zevitz, 2006).
Given the rather unique societal disdain directed toward sex offenders, it is unsurprising that a robust body of literature has examined public perceptions of sex offenders and punishment (Brown et al., 2008; Craun & Theriot, 2009; Kernsmith et al., 2009; Levenson et al., 2007; Schiavone & Jeglic, 2009). Relatively few studies, however, have endeavored to examine perceptions of the death penalty in relation to sexual offending. A 2008 Quinnipiac poll reported that 47% of respondents supported the death penalty in cases of rape, while 65% supported the death penalty in cases involving child victims. Subsequent work by Mears et al. (2008), Mancini and Mears (2010), and Wiersma and Siedschlaw (2016) yielded similar findings. In an effort to explain disparate levels of support for the death penalty in sex offenses involving child versus adult victims, Mancini and Mears (2010) examined demographic and attitudinal data from 1,101 respondents who participated in a telephone poll conducted by the Minneapolis Star Tribune in 1991. They noted markedly higher levels of support for the death penalty in cases involving child (51%) versus adult victims (27%). Interestingly, however, respondent gender, race, and political affiliation—factors that are frequently correlated with general support for the death penalty (e.g., Applegate et al., 2002; Oliphant, 2018; Unnever & Cullen, 2005)—failed to predict support for the death penalty in sex offenses involving child or adult victims. In contrast, respondent levels of education and their perceptions of sex offender recidivism and failures of the court system were consistent predictors across both models.
The Current Study
Despite the prevalence of research on attitudes toward the death penalty and its application to cases involving sex offenses, no known research has tested support for the death penalty in cases of sexual assault across victim age categories. While research has explored differences across dichotomous measures of child versus adult victims (e.g., Mancini & Mears, 2010), questions remain regarding how victim age relates to death penalty support. The purpose of this study is to examine support for the death penalty in cases of sexual assault across victim age categories using a sample of college students at a university in the Southeastern U.S.
Data, Measures, and Methods
Data
The data used in this study were collected from a mid-size university in the Southeastern United States during the Spring 2019 semester and received ‘exempt designation’ from the university’s IRB committee (IRB# 19-045). At the time of the study, a total of 10,176 undergraduate students were enrolled. Initially, the research team selected a stratified random sampling technique whereby the strata of the sampling frame were defined by department and course level (e.g., 1000–4000), and classes selected from within each stratum. This approach proved problematic, however, as a consequence of scheduling conflicts and several non-responses from professors of selected courses. As a result, this study adopted a convenience sampling technique that allowed for the sampling of 16 different courses, largely from the social sciences. These included five 1000-level, three 2000-level, six 3000-level, and two 4000-level courses nested within six academic programs. Researchers met with students during their scheduled class times, explained the purpose and voluntary nature of the study, and distributed consent forms with each survey. The survey consisted of 78 items, including demographic information, that corresponded with perceptions of a range of criminal justice issues. A total of 974 paper-pen surveys instruments were delivered to the selected courses based on enrollment records maintained by the university. Of this number, 451 did not participate as a result of either prior completion of the survey, absenteeism, or refusal. A total of 523 surveys were completed by the students. Of these, 49 were not usable due to incomplete/missing responses, resulting in a final sample of 474.
We recognize that college students are a somewhat unique population, which might generate concerns related to generalizability. Indeed, college campuses tend to be diverse environments that provide greater opportunities for contact between races, ethnicities, and cultures (Mancini et al., 2012; Rankin & Reason, 2005). As noted by Wiecko (2010), however, college students tend to exhibit attitudes toward serious offenses and criminalistic behavior that closely mirror those of the general public. This assertion finds support in a number of perception-based studies (e.g, Mancini et al., 2012; Piquero & Bouffard, 2007), including those of capital punishment (e.g., Burgason & Pazzani, 2014; Wiersma & Siedschlaw, 2016), which provides a layer of support for the generalizability of the current study.
Dependent Variables
A total of six dependent variables are included in our analyses. Respondents were presented with a question asking, “For which of the following crimes do you believe the death penalty should be an option as a sentence? (check all that apply).” This question was followed by five options related to sexual offending: (1) Rape/sexual assault of an adult person (age 18 and over); (2) Rape/sexual assault of a child (age 15–17); (3) Rape/sexual assault of a child (age 12–14); (4) Rape/sexual assault of a child (age 9–11); and (5) Rape/sexual assault of a child (ages 8 and under). 1 Each was coded as a dichotomous measure based on the binary nature of available responses (1 = yes; 0 = no). The sixth dependent variable is a summed scale of item responses for questions 1 through 5, whereby larger values represent support for the death penalty across an increased number of victim age categories and, in turn, stronger support for the death penalty as sentencing option in cases of rape and sexual assault (α = .966).
Independent Variables
The purpose of this study was to explore support for the death penalty as sentencing option in cases of rape/sexual assault across victim age categories. As noted in the review of the literature, respondent characteristics ranging from biological sex to political orientation may influence support for the death penalty (e.g., Applegate et al., 2002; Oliphant, 2018; Messner et al., 2006; Unnever & Cullen, 2010). Interestingly, however, the work of Mancini and Mears (2010) suggests that demographic factors may exhibit comparatively little influence over support the death penalty in cases of sexual victimization of adults and children. To further probe this question, a number of independent variables were included in our analyses. Biological sex (1 = male, 0 = female), race (1 = White, 0 = non-White), political affiliation (1 = Republican, 0 = other), and college major (1 = Criminal Justice, 0 = other) were each operationalized as dichotomous measures. 2 Parents’ education was treated as a proxy of socio-economic status and operationalized through a question asking respondents to indicate their parents’ highest level of education (1 = bachelor’s degree or higher, 0 = other), while semester standing was measured through an item asking respondents to indicate whether they were a Freshman (1), Sophomore (2), Junior (3), or Senior (reference category). Fear of victimization was measured by a multiple-item index (Cronbach’s α = .89) frequently used in prior works (e.g., Ferraro, 1995; Jang et al., 2015; Warr, 2000). Respondents were asked to “indicate how afraid you are of the following things happening to you” using a 5-point Likert-style scale ranging from 1 (not afraid at all) to 5 (very afraid). Victimization categories included (a) being murdered, (b) being robbed or mugged, (c) being attacked by someone with a weapon, (d), being sexually assaulted, (e) being stalked, (f) having things stolen from your home/apartment, and (g) having your car stolen.
Analytic Strategy and Hypotheses
In accordance with extant literature, several hypotheses are tested. As noted in the review of the literature, multiple studies have described comparatively higher levels of support for the death penalty among males (e.g., Applegate et al., 2002; Oliphant, 2018). Consistent with these findings, we hypothesize that:
Hypotheses 1: Identifying as male will be associated with an odds-increase in support for the death penalty in cases of sexual assault regardless of victim age.
Similarly, prior works have consistently demonstrated that political conservatism (e.g., Messner et al., 2006; Unnever & Cullen, 2010) and majoring in criminal justice (e.g., Farnworth et al., 1998; Lambert et al., 2008) are positively correlated with support for the death penalty. In accordance with these works, it is hypothesized that:
Hypothesis 2: Identifying as Republican will be associated with an odds-increase in support for the death penalty in cases of sexual assault regardless of victim age.
Hypothesis 3: Being a Criminal Justice major will be associated with an odds-increase in support for the death penalty in cases of sexual assault regardless of victim age.
Further, and consistent with the results of studies by Costelloe et al. (2009), Unnever and Cullen (2007), and Wu et al. (2011), we expect that:
Hypothesis 4: Indicating higher levels of fear of crime will be associated with an odds-increase in support for the death penalty in cases of sexual assault regardless of victim age.
In contrast, based on the work of Costelloe et al. (2009), we anticipate that:
Hypothesis 5: Having a parent who graduated from college will be associated with an odds-decrease in support for the death penalty in cases of sexual assault regardless of victim age.
Given the purpose of the study and dichotomous nature of the dependent variables representing distinct victim age categories, binary logistic regression was selected as the most appropriate statistical approach. Furthermore, we applied equality of coefficients tests described by Paternoster et al. (1998) 3 in order to determine if differences between victim age categories in terms of the influence of individual independent variables achieved statistical significance. The ordinal nature of the scaled summary measure required application of ordinal regression.
Results
Descriptive Statistics
Descriptive statistics are displayed in Table 1. A trend emerged in terms of support for the death penalty as a sentencing option in cases of rape/sexual assault, with support increasing as victim age decreased. The lowest level of support for the death penalty involved adult victims of rape/sexual assault (24%), while the highest levels of support were for cases involving child victims aged eight and under (39%). Overall, approximately 60% of the sample did not support the death penalty as a sentencing option for any victim age category, as evidence by the modal response of zero for the summary scale. On average, however, respondents indicated low to moderate levels of support for the death penalty as a sentencing option in cases of rape and sexual assault (M = 1.67).
Descriptive Statistics for Variables in Analysis (n = 474).
The sample was 33% male and 77% White, which aligned relatively well with the demographics of the sample university. 4 Indeed, at the time of survey distribution the university maintained a student population that was 44% male and 75% White. Further, 27.6% of the students were freshmen, 20.3% were sophomores, 22.2% were juniors, and 28.9% were seniors—figures that also align relatively well with our sample characteristics. Thirty-two percent of respondents identified as Republican, while 61% indicated that at least one of their parents had earned a bachelor’s degree or higher. Nearly one-third of the respondents were majoring in criminal justice and a desirable degree of variation was observed in semester standing. The sample exhibited moderate-level fear of victimization, with an average of 22.85 on the 35-point scale.
Regression Analysis
Table 2 describes the results of the six regression models included in our analyses. In Model 1, ordinal regression was used to examine the influence of the independent variables on the additive index denoting support for the death penalty across all victim age categories. Model fit statistics were appropriate and, as additional variables were added, -2 Log Likelihood became smaller than the constant-only model which is indicative of model improvement. Biological sex, political orientation, college major, fear of crime, and parents’ level of education each shared statistically significant relations with the outcome measure. Specifically, respondents who were male, Republican, majoring in criminal justice, and maintained comparatively higher levels of fear of crime exhibited increased levels of support for the death penalty as a sentencing option in cases of rape and sexual (e.g., comparatively higher scores on the additive scale). In contrast, students whose parents were more highly educated, our proxy for socio-economic status, exhibited comparatively lower levels of support for the death penalty in cases of rape and sexual assault (e.g., lower scores on the additive scale).
Regression Analyses Predicting Support for Death Penalty in Rape/Sexual Assault.
Note. Standard errors in parentheses. Model 1: Ordinal logistic regression; Models 2-6: Binary logistic regression.
p ≤ .05. **p ≤ .01. ***p ≤ .001.
In order to explore if these findings were robust to disaggregation by victim age, Models 2 through 6 used binary logistic regression to examine levels of support for the death penalty in cases of rape and sexual assault using distinct victim age categories. In each case, model fit statistics were appropriate and, as additional variables were added, -2 Log Likelihood became smaller than the constant-only model—indicating model improvement. In support of Hypothesis 1, male respondents were more likely to support the death penalty as a sentencing option in cases of rape and sexual assault across all victim age categories involving those victims under the age of 18. The relation between respondent sex and support for the death penalty was reduced to marginal statistical significance in cases of rape and sexual assault involving adult victims (p = .059). However, application of equality of coefficients tests revealed no statistically significant differences across 10 model comparisons (see Table 3). This finding suggests that the impact of respondent sex is salient across victim age categories.
Comparison of Coefficients Tests for Models 2-6 (z Value Displayed).
A similar trend was observed in terms of political affiliation. Respondents who self-identified as Republican were more likely to support the death penalty in cases of rape and sexual assault—findings that yielded strong support for Hypothesis 2. The lone exception was cases of rape and sexual assault involving victims between the ages of 15 and 17, though this relation did reach marginal significance (p = .093). Once again, however, application of equality of coefficients tests revealed no statistically significant differences across 10 model comparisons. This finding provided an additional layer of support for Hypothesis 2.
Prior works (e.g., Farnworth et al., 1998; Lambert et al., 2008) have suggested that college students majoring in criminal justice exhibit comparatively higher levels of support for capital punishment, prompting the current study to hypothesize a similar effect. Although criminal justice majors were more likely than their peers to support the death penalty as a sentencing option in cases of rape and sexual assault involving victims under the age of 12, this effect reached only marginal significance in cases involving victims between the ages of 12 to 14 (p = .060) and 15 to 17 (p = .053), and non-significance in cases involving adult victims (p = .502). There were, however, no statistically significant differences between coefficients across the 10 model comparisons—providing a greater level of support for Hypothesis 3.
In accordance with Hypothesis 4, the influence of fear of crime was highly consistent. Respondents who maintained higher levels of fear of crime were more likely to support the death penalty as a sentencing option in cases of rape and sexual assault. Moreover, the positive influence of fear of crime was salient across all victim age categories. Once again, these findings were robust in equality of coefficients tests, with no statistically significant differences emerging in any of the 10 model comparisons.
Our findings were generally supportive of prior research (e.g., Costelloe et al., 2009) that has described a negative correlation between socio-economic status and support for the death penalty. Parents’ level of education, our proxy for socio-economic status, was associated with decreased odds of supporting the death penalty as a sentencing option in cases of rape and sexual assault across all victim age categories involving children. The negative relationship between parents’ level of education and support for the death penalty extended to cases of rape and sexual assault involving adults, but reached only marginal significance (p = .095). The lack of statistically significant differences between all 10 model comparisons provided additional support for Hypothesis 5.
Finally, in terms of controls variables, semester standing appeared to exert little influence over support for capital punishment in cases of rape and sexual assault. In comparison to seniors, freshmen were more likely to support the death penalty as a sentencing option in cases involving victims between the ages of 15 and 17, as well as those aged 8 and under. Respondent race, however, was not associated with the odds of supporting the death penalty in cases of rape and sexual assault. Equality of coefficients tests revealed no statistically significant differences between coefficients for any of the controls in any model comparison.
Discussion and Conclusions
The purpose of this study was to examine support for the death penalty as a sentencing option in cases of rape/sexual assault across victim age categories. Prior works have explored levels of support for capital punishment, and the factors that influence those perceptions, in cases of sexual offending against adult versus child victims but have used a simple dichotomy to separate the two (e.g., Mancini & Mears, 2010; Mears et al., 2008; Quinnipiac University, 2008). This approach leaves questions concerning how levels of support might vary across multiple age categories. Specifically, prior research appears to have implicitly assumed that support for the death penalty in cases of child victimization is uniformly higher in comparison to cases involving adult victims. Our findings suggest this may not be the case.
Consistent with the trend discussed in the extant literature, we found that support for the death penalty as a sentencing option was lowest in cases of adult victims (24%) and highest in cases of child victims aged 8 and under (39%). Importantly, however, support for the death penalty in cases of child victimization was not consistent across victim age. Instead, support for the death penalty decreased as victim age increased. Thus, while our findings are supportive of the assertion by Mancini and Mears (2010) that agreement concerning capital punishment is greater when the focus is on child victims of sexual assault, they also suggest that such agreement is not uniform. Further, this lack of uniformity suggests that examination of perceptions of capital punishment in relation to sexual offending remains a fruitful line of inquiry.
Beyond our descriptive statistics, the results of multivariate analyses merit further discussion. As described in the review of the literature, respondent sex, race, and political orientation tend to exhibit rather robust associations with attitudes toward capital punishment (Applegate et al., 2002; Oliphant, 2018; Unnever & Cullen, 2005). In contrast, Mancini and Mears (2010) noted that these factors were not correlated with variation in levels of support for the death penalty in cases of sexual offending involving either child or adult victims. Our findings as they relate to respondent sex and political affiliation are, therefore, more consistent with the general literature on death penalty attitudes. This result provides a degree of support for the use of our sample, as well as the argument by Wiecko (2010) and others that the opinions of college students tend to mirror those of the general public when it comes to serious crime. Equally important, the positive influence of biological sex and political affiliation was rather uniform.
The relation between biological sex and support for the death penalty as a sentencing option was, as expected, positive and statistically significant across all models involving child victims. This relationship, however, was reduced to marginal significance in cases involving adult survivors of rape and sexual assault. At face value, this finding might suggest that males perceive sexual victimization of adults as less serious in comparison to children. This notion would further invoke issues related to perceptions of “legitimate” rape/sexual assault and victim blaming based on credibility, behavior, and appearance that are common in cases involving adult survivors (Abbey & Jacques-Tiura, 2011; Basow & Minieri, 2011; Burt, 1980; Edwards et al., 2011; Payne et al., 1999; Suarez & Gadalla, 2010). However, the results of equality of coefficients tests clearly indicate this is not the case. Not only does this suggest the robustness of biological sex as a predictor of support for the death penalty in cases of rape and sexual assault, but this finding is consistent with studies of general support for capital punishment (e.g., Applegate et al., 2002; Oliphant, 2018).
A similar theme emerged in relation to the impact of political affiliation. As expected, conservative political ideology consistently exhibited a positive correlation with support for the death penalty as a sentencing option in cases of rape and sexual assault. This association was, however, reduced to marginal significance in cases where the victim was between the ages of 15 and 17. Once again, however, application of z-tests to the coefficients for political affiliation revealed that the influence of conservative political ideology was consistent between models. Overall, this finding is consistent with prior studies that have examined the relation between political conservatism and support for the death penalty (e.g., Bobo & Johnson, 2004; Messner et al., 2006; Wozniak & Lewis, 2010).
Like male sex and affiliation with the Republican party, it was expected that criminal justice majors would exhibit relatively higher levels of support for the death penalty as a sentencing option in cases of rape and sexual assault across all victim age categories. This hypothesis was derived from the work of Farnworth et al. (1998) and Lambert et al. (2008), each of which suggested that criminal justice majors were more supportive of the death penalty in comparison to non-criminal justice majors. Our logistic models provided only partial support for this hypothesis, as criminal justice majors appeared comparatively more supportive of the death penalty as a sentencing option only in cases where the victim was age 11 or younger. Absent the null results of z-tests, this finding would have suggested that criminal justice majors assume a more punitive stance on sex crimes involving victims who are more likely to be pre-pubescent, less likely to be sexually active, and less able to resist their offenders. Instead, the influence of majoring in criminal justice is a robust correlate of support for the death penalty in cases of rape and sexual assault, a finding in line with the works of Farnworth et al. (1998) and Lambert et al. (2008).
The impact of fear of crime and parents’ level of education, our proxy for socio-economic status, were far less nuanced across logistic models. Across all victim age categories, fear of crime shared a positive and statistically significant correlation with support for the death penalty as a sentencing option. This finding is not altogether surprising given the perpetual panic associated with sexual offending (Burchfield et al., 2014) and the consistent relation between fear of crime and support for aggressive crime control strategies, including the death penalty, described in the literature (Costelloe et al., 2009; Unnever & Cullen, 2007; Wu et al., 2011). However, it does demonstrate that the impact of fear of crime is salient between samples of college students and the general public (e.g., Wiecko, 2010; Wu et al., 2011). Similarly, respondents’ whose parents were college educated were significantly less likely to voice support for the death penalty as a sentencing option across all victim age categories with the exception of those who were 18 or older. We do not interpret the latter to suggest increasing levels of support for the death penalty among this demographic, but rather an overall attenuation in levels of support for the death penalty in cases of sexual assault involving adult victims. Indeed, z-tests revealed that the differences between log-odds coefficients for parental education were statistically non-significant. Moreover, the overall trend observed among respondents’ whose parents are college educated is consistent with prior works that have explored the relation between socio-economic status and support for the death penalty (Costelloe et al., 2009).
In contrast, our findings regarding respondent race represent a significant departure from the literature as it relates to capital punishment, in general. A multitude of studies have suggested that race is a robust predictor of support for the death penalty, with Whites exhibiting higher levels of support in comparison to non-Whites (e.g., Barkan & Cohn, 1994; Cochran & Chamlin, 2006; Oliphant, 2018). Our results, however, indicate that race may not factor into support for the death penalty in relation to rape and sexual assault. This finding is, however, consistent with the work of Mancini and Mears (2010), who relied on a sample of the general public.
Similarly, the work of Bohm et al. (1991) suggests that support for the death penalty diminishes with increased education on issues related to the death penalty. Our findings provide limited support for this assertion. Overall, levels of support for the death penalty as a sentencing option for any age category were noticeably lower in comparison to prior studies that have explored support for the death penalty in general (e.g., Dotson & Carter, 2012; Oliphant, 2018) or as it relates specifically to sexual offending (e.g., Mancini & Mears, 2010; Quinnipiac University, 2008)—a finding that, in consideration of the work of Bohm et al. (1991) and Fox et al. (1991), should be expected among a sample of college students. Further, prior studies suggest that young adults are less supportive of the death penalty than middle-aged and older adults (e.g., Anderson et al., 2017; Oliphant, 2018). In general, however, semester standing failed to exhibit strong or consistent influence over support for the death penalty as a sentencing option.
Despite its contribution to the literature, this study does suffer a number of limitations that should be discussed. Although there is a degree of face-validity in our interpretation of the relationships between the variables we examined, the cross-sectional nature of our data does limit inference of causality. Second, we must remind the reader of issues of representativeness related to our sample. As noted in our discussion, some of our findings may well be influenced by the fact that they were drawn from a sample of college students rather than the general population. Moreover, our sample is drawn from a university in the Southern U.S., a region that tends to maintain comparatively higher levels of support for capital punishment (Baumer et al., 2003; Barkan & Cohn, 2010; Borg, 1997). However, given the extent to which the relationships described here are consistent with prior studies of support for the death penalty among the general public, our findings lend support to the argument by Wiecko (2010), that perceptions of colleges students in relation to serious crime closely resemble those of the general public.
Finally, our study assuredly suffers from omitted variable bias. Prior works have suggested that factors such as religious beliefs (Bader et al., 2010; Rade et al., 2017; Unnever et al., 2006) and negative perceptions of the criminal justice system (Tyler & Weber, 1982; Unnever & Cullen, 2007) may influence levels of support for capital punishment in general, while perceived ineffectiveness of the courts and high recidivism rates among sexual offenders (Mancini & Mears, 2010) may influence support for application of the death penalty in sex crimes. Whether these relationships hold across multiple victim age categories remains an open empirical question—one we encourage future studies to explore. It is also possible that respondent support for the death penalty as a sentencing option in cases of rape and sexual assault may be influenced by situational characteristics that were not included in our survey, such as victim injury, victim and offender sex, and weapon use by the assailant. In this vein, future studies of this topic would likely benefit from the use of vignettes similar to those adopted by researchers examining support for the death penalty in cases of murder (e.g., Burgason & Pazzani, 2014) in order to determine the degree to which the findings of Cochran et al. (2019), Durham et al. (1996), and Jennings et al. (2017) extend to rape and sexual assault.
Mancini and Mears (2010) have argued that the degree to which support for the death penalty diverges across different types of offenders represents an important avenue of inquiry. This study is an effort to answer that call. There are also important implications for policy and practice that stem from such exploration. In the U.S., perpetual panic associated with sex crimes has promoted increasingly harsh punishments meted out against sex offenders (Burchfield et al., 2014; Calkins et al., 2014; Dierenfeldt & Carson, 2017; Vasquez et al., 2008). This trend culminated with the decision by multiple state governments to allow capital punishment as a sentencing option in cases involving child victims before being struck down in Kennedy v. Louisiana in 2008.
The 5-4 decision in Kennedy was, however, a close one. Prior works have consistently revealed moderate to high levels of support for the death penalty in cases of sexual victimization, particularly those involving children (e.g., Mancini & Mears, 2010; Mears et al., 2008; Quinnipiac University, 2008). Consistent with this trend, a large proportion of our respondents supported the death penalty as a sentencing option in cases of rape and sexual assault involving children—including nearly 40% who supported the death penalty if the victim was age 8 or younger. Further, despite their general ineffectiveness, harsh sex offender laws enjoy broad public support (e.g., Jenkins, 1998; Meloy et al., 2013). If such high levels of support remain, it may only be a matter of time before states reexamine the political and legal viability of the death penalty in cases of sexual offending. A more conservative Supreme Court might also be more sympathetic to such legislation. Indeed, in Justice Alito’s dissenting opinion, he argued that the Court overstepped in its reversal of both the Kennedy and Coker decisions, and that states should be allowed to maintain such laws in accordance with the will of their respective populations and in accordance with evolving social standards of morality and harm. He further asserted that “in the eyes of ordinary Americans, the very worst child rapists. . .are the very epitome of moral depravity” (Kennedy v. Louisiana, 554 U.S. 2008), while suggesting that, in certain sex offenses involving child victims, capital punishment was proportional to the harm and moral depravity of the offense. Importantly, he was joined in his dissent by Chief Justice Roberts, Justice Thomas, and Justice Scalia. Following the passing of Justice Scalia, the three remaining dissenters in Kennedy have since been joined by conservative Justices Gorsuch, Kavannaugh, and Barrett. It seems prudent, then, to take measures to examine not only perceptions of the death penalty in relation to sex crimes, but the factors that influence those perceptions—particularly among demographics that have traditionally voiced lower levels of support for capital punishment (e.g., college students).
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
