Searching for a Comprehensive Understanding of Collegiate Sexual Assault Rates:  Assessing Cross-Campus Variance of Sexual Assault Rates According to Community, Institutional, and Student Characteristics

Steven Keener, M.S.

Abstract: College sexual violence is an issue that has recently received a growing amount of attention. Despite this attention, researchers have yet to take a comprehensive approach that utilizes student, institutional and community level factors to better understand the variance of sexual victimization rates across colleges and universities. This study first sets the foundation for this type of research by identifying a theoretical framework that can guide studies of this nature. The study then works through a number of related empirical studies that help give credence to the necessity of this research. An initial regression analysis is included to set the stage for future studies. The study concludes with the policy implications and how this research will impact the work of administrators and policymakers that are seeking to reduce campus sexual victimization rates. 

Introduction

According to End Rape on Campus, an advocacy organization for survivors of sexual violence, twenty-eight out of the top fifty "best universities" in U.S. News and World report are under federal investigation for their handling of sexual assault (United States Department of Education, 2015).

In the last decade, research has indicated the staggering presence of sexual assault on college campuses: approximately 20 percent of women and 6 percent of men are sexually victimized during their college careers. These numbers have pushed the issue of campus sexual assault, arguably a millennial driven issue, given the evolution of the college experience in the 21st century, to the forefront of the national policy agenda (Fisher, Cullen, & Turner, 2000; Krebs et al., 2007; “Not Alone”, 2014; Sampson, 2002; Sloan & Fisher, 2010). 

Some studies suggest that approximately 95 percent of sexual assault victims on college campuses fail to report the incident to law enforcement or campus officials, suggesting that the high rates already reported fail to reflect the true prevalence of the problem (Karjane, Fisher & Cullen, 2005; Fisher et al., 2003). The physical and emotional consequences of sexual assault, especially when unreported, are wide-ranging; 34 percent of college sexual assault survivors reportedly experience Post Traumatic Stress Disorder (PTSD) after the incident and 33 percent experience depression (both are at higher rates than experienced by Iraq and Afghanistan War Veterans [i]). More, researchers have estimated that the economic cost of assault ranges from $87,000 to over $240,000 due to medical treatment, counseling and other negative impacts to quality of life (Kilpatrick et. Al, 2007; “Not Alone”, 2014).

This type of data is starting to receive widespread attention from the general public and policymakers alike. Work by activists like Emma Sulkowicz, a Columbia student who in 2014 carried the mattress on which she was sexually assaulted to all of her classes, and advocacy organizations, like End Rape on Campus, have brought the issue to national attention. Since then, public information campaigns, such as President Obama's It's On Us campaign, and federal legislation, such as Senator Claire McCaskill’s Campus Accountability and Safety Act, have begun to tackle the issues [ii].

Despite the current steps being taken to confront sexual violence on college campuses, research into the issue has not caught up. Particularly, understanding what drives officially reported sexual offense rates across academic institutions is a key body of knowledge that research has yet to fully tackle. Inherent differences between schools, such as location and student population, impact both the number of sexual offenses that occur on-campus and the rate at which these offenses are officially reported; but what breeds both of these rates across institutions is lesser understood.

While academics have studied the impact of disaggregated factors on campus sexual assault rates, a lack of data has prevented researchers from conducting a comprehensive assessment of how and why different institutions have different rates of sexual assault. This paper attempts to lay the groundwork for a better understanding of the relationship between a college’s community, institutional and student characteristics and their officially reported sexual offenses. To answer this question, this paper first points to Schwartz and Pitta’s Feminist Routine Activities Theory as a simple, yet comprehensive theoretical framework to guide studies of this nature. It then reviews the literature, beginning with studies that have analyzed multi-level characteristics of academic institutions and their general campus crime rates, and then delving into studies that have analyzed specific correlates of campus sexual assault rates. With a contextual research foundation in place, the early framework for a comprehensive analysis is proposed, one that examines a sample of colleges and their student, community and institutional characteristics to see which factors most strongly influence rates of sexual assault. The paper concludes by postulating the policy implications that arise from these elementary findings and the ways new-age tools might help pave the way for similar analyses in the future.

Theoretical Framework

The primary theoretical framework used in campus crime studies is Cohen and Felson’s (1979) Routine Activities Theory, which posits that most crime, in the absence of ‘capable guardians’, requires the convergence of likely offenders and suitable targets. While this theory has a solid foundation, it fails to consider offender motivations (DeKeseredy et al., 1995). Schwartz and Pitta’s (1995) Feminist Routine Activities Theory addresses this weakness, arguing that common offenders of campus sexual assault are those from male peer groups that perpetuate the sexual exploitation of women. Theoretically, these offenders are stimulated by an existing culture and societal messages that enable men to physically and sexually overpower women with few social consequences. This study operates under the assumptions present in both theories, that potential male offenders of sexual assault, both in and outside of the student body, come from male peer groups and seek suitable targets, college women, with an absence of capable guardians. The value of this theoretical lens is its guidance of what factors might comprise a comprehensive analysis – it informs researchers to seek out characteristics that either measure the relevant factors (common offenders, suitable targets, capable guardians) or characteristics that may influence the presence of these factors on or around a college campus. For example, the proportion of male students or the level of on-campus security at a college or university would indicate the presence of potential male offenders present on campus and the provided level of protection against these offenders. 

Literature Review

Prevalence of Campus Sexual Assault

It is estimated that approximately 1 in 5 women and 1 in 17 men will be sexually victimized during their time in college (Fisher, Cullen, & Turner, 2000; Krebs et al., 2007; “Not Alone”, 2014; Sampson, 2002; Sloan & Fisher, 2010). In addition, 13 percent of college women report being stalked while in college. Yet, approximately 95 percent of collegiate sexual assault victims do not report their victimization through official avenues, such as law enforcement or campus officials (Fisher, Cullen & Turner, 2000; Karjane, Fisher & Cullen, 2005; Fisher et al., 2003). Research also suggests that, more broadly, sexual assault is the most underreported violent crime in the United States (Fisher et al. 2010; Fisher et al., 2003; Lizotte, 1985; Reddington & Kreisel, 2009; Russell, 1982; Sampson, 2002). 

Characteristics Associated with Campus Crime

Past studies have analyzed general campus crime rate by examining academic institutions’ community or location, student body and institutional characteristics. Community characteristics describe the city or county in which an institution is located, informing potential insight on likely offenders, victims or capable guardians. For example, cities with large populations have more people present that could fill these roles. Which characteristics of a college’s location might generate offenders? Who might act as a guardian for victims? Research has identified smaller relationships – such as a correlation between an area’s unemployment rate and a college’s campus crime rate – but has yet to tackle a broader assessment of an academic institution’s surrounding area (Fox & Hellman, 1985; McPheters, 1978).

Student characteristics include the makeup, features and details of the student body, all of which might illustrate potential victims or offenders, and also point to contextual factors that might increase assault rates (Volkwein et al., 1995). Past research suggests that predictors of higher campus crime rates include more students living on-campus, more minority students and more full-time students, (McPheters, 1978; Fernandez & Lizotte, 1995). Nobles et al. (2013) used geographic information systems (GIS) mapping technology to examine crimes occurring both within a school’s borders and nearby. They found that males were more likely to be arrested on campus than females, and in particular, white males were more likely than males of other races. This type of research evidences the importance of student characteristics, those that describe the makeup of student body both on an individual and aggregate level, in illuminating what drives sexual assault on college campuses.

Institutional characteristics describe aspects of a college or university outside of its student body, such as tuition costs, type of degrees awarded, funds spent on security or gender ratio of faculty. These characteristics illustrate an institution’s academic and organizational infrastructure, that which creates an environment for likely offenders, victims or guardians. (Volkwein et al., 1995). For example, an institution’s student-to-faculty ratio could depict the presence of capable guardians on-campus. More, McPhethers (1978) identified a relationship between higher campus crime rates and a larger amount of funding spent on campus security – while schools with higher campus crime rates are likely to then spend more money on security, the nuances of this relationship are still valuable to explore. For example, when controlling for an array of other institutional, student or community factors, does more money spent on campus security still lead to higher campus crime rates? Do dollars spent on campus security not lead to lower rates? Other significant characteristics that are associated with increased campus crime rates include larger campuses, higher enrollment totals, higher cost of room and board and increased institutional wealth (Bromley, 1995; Fox & Hellman, 1985; Fernandez & Lizotte, 1995; Volkwein et al., 1995).

Individual and Group Correlates of Sexual Assaults

Most studies that explore the relationship between campus sexual assault rates and specific characteristics have focused on how individual or group-level indicators singularly influence sexual assault rates. Within individual-level assessments, research focuses primarily on victim intoxication. Mohler-Kuo et al. (2004) used data from 119 schools to demonstrate that approximately 72 percent of rape victims suffered the incident while intoxicated. They also discovered that white women were more likely to have been raped while intoxicated, but less likely to suffer an attack when there was no alcohol involved compared to women of other races. Students residing on campus were 1.4 times more likely to experience a rape incident, and students residing in sorority houses were almost 3 times more likely. Heavy episodic drinking and drug use were strong predictors of rape – college women who reported frequent and occasional episodes of heavy drinking were 7.8 times more likely to experience a rape incident while intoxicated and women that reported using illicit drugs were 4.6 times more likely to experience a rape incident while intoxicated compared to women who did not use drugs (Mohler-Kuo et al., 2004). These findings, while valuable in understanding what situations and demographics lead to increased risk, focus primarily on individual experiences and fail to uncover what other characteristics were present across the general community or academic institution.

In line with the Feminist Routine Activities Theory, group-level assessments have focused on the impact of peer groups on increased risk of sexual assault. Martin and Hummer (1989) identify factors that contribute to a fraternity’s level of objectification of women, including how much members compete with members of rival fraternities over sports and women, and might also contribute to increased sexual assault rates. Other characteristics of fraternities that link to objectification of women include common features of masculinity, preoccupation with loyalty and group secrecy (Martin & Hummer, 1989). Regarding likelihood to sexually exploit college women, fraternities have been classified as high- and low-risk, with high-risk fraternities more likely to perpetuate misogynistic attitudes and behaviors. Men in high-risk fraternities are also more likely to degrade women or treat them as subordinates, dislike committed relationships and engage more in the ‘hook up’ culture (Boswell & Spade, 1996; Humphrey & Kahn, 2000). Another study found that the women most likely to be sexually victimized were those that are not previously known to fraternity members, as the nameless and faceless nature of these women render them easier to forget (Boswell & Spade, 1996).

Institutional Level Correlates of Sexual Assaults

The rules, practices and distribution of resources of an institution can perpetuate gender inequality; as these characteristics and interactions can be gendered, they have the ability to influence interactions between men and women on campus and increase risk of sexual assault. Culture develops in response to institutional arrangements, suggesting that a change in culture first requires the acknowledgement of relevant institutional factors (Armstrong et al., 2006; Swidler, 2001). Armstrong et al. (2006) argued that, based on a collection of research, we should expect campuses with lower alcohol use, minimal party scene and more racial diversity to have lower sexual assault rates.

Relationship Between Community Variables and Interpersonal Violence (IPV)

While there’s little research on the relationship between campus sexual assault rates and features of the surrounding community, researchers have analyzed links between community-level variables and violence, or more specifically interpersonal violence (IPV), in the general population. Significant predictors of increased IPV levels include neighborhood disorder, unemployment rates, low levels of education, residential instability and pre-existing community violence (Caetano et al., 2010; Cunradi, 2007; McKinney et al., 2009; Obasaju et al., 2009; Waller et al., 2011; Benson et al., 2003; Browning, 2002; Li et al., 2010; Raghaven et al., 2009; Reed et al., 2009). While urbanicity has been found to play a role in increased IPV rates, research on this relationship is mixed (Antai, 2001; Boyle et al., 2009). Similar to the argument made for institutional characteristics, the contextual insight gained by also examining community-level characteristics can lead to a clearer understanding of what about a certain environment or area leads to increased risk of violence or crime.

Initial Analysis

Methodology

Research has yet to produce a comprehensive study analyzing the relationship between colleges’ sexual victimization rates and the combination of its community, student and institutional characteristics. Using the totality of the literature as guidance, an initial analysis is provided here to illustrate the direction in which future research can and should move if we seek to better understand the drivers of campus sexual assault. However, the analysis suffers from obvious methodological weaknesses, most of which surround the dependent variable, sexual offenses, and the availability of necessary data.

The data on sexual offenses for each school in the sample comes from those sexual offenses that occurred per 100,000 students at that given institution and were reported to official entities, such as campus law enforcement or school administrators. While availability of data more broadly differed by institution, data for officially reported sexual offenses is federally mandated under the Clery Act.[iii] However, this data does not capture unreported sexual assaults. This points to a gap in information that policymakers and institutional actors should address, as the lack of comprehensive data in this analysis elucidates a key need for future studies of this nature. In addition to a lack of necessary data, the analysis also suffers from the absence of desired data, including information on-campus law enforcement, counseling, Greek life and alcohol use. These variables represent potentially important predictors of sexual victimization rates that ideally should be included in future analyses. When researchers have access to data of this sort, they can enhance the preliminary model detailed below, using their findings to examine influence of more granular contextual factors.

This preliminary analysis employs a simple cross-sectional design, using a regression model to frame community, student and institutional characteristics as predictor (independent) variables, and officially reported sexual offenses as the criterion (dependent) variable. The sample includes a random selection of 286 U.S. higher education institutions that fulfill the following conditions: 2-year or above, non-profit, with a physical campus and at least 1,000 students.[iv] To include the relevant collection of factors, data was collected across various sources from the 2010-2011 academic year, as this was the most up-to-date publicly available data for all variables.

Using the guidance of both prior research and the Routine Activities theoretical framework, over 100 potential independent variables were selected. Decisions to exclude certain characteristics were first made based on theoretical decisions, such as if they have been given credence in prior relevant research or whether they fit in one of three tenants of the Feminist Routine Activities theory (motivated male offenders, suitable female targets, capable guardians). Certain characteristics were also excluded for empirical reasons, such as a lack of variance in the data.[v] Table 1 details the variables used in the models and their sources

Results

Community, institutional and student characteristics exist across all colleges and universities, rendering each of these categories an important addition to an analytical model that attempts to reflect the true nature and experience of academic institutions. Similar to the work of Volkwein et al. (1995), community characteristics were added first to the model, as they arguably represent factors that craft the foundation or infrastructure of an academic institution. This ordering was also supported by their early analysis that found certain crime rates were consistently higher off campus – specifically, students were safer from violent crime when they were on-campus rather than off-campus, and that property crime rates were substantially higher in the surrounding community than on corresponding campuses (Volkwein et al., 1995). Accordingly, Volkwein et al. (1995) used community characteristics as the foundation of their model due to the belief that campuses are more likely to attract crime from their surroundings than engender it themselves, and that motivated offenders are more prevalent off campus than on it.

 For the first combined model, student characteristics were added next, then institutional characteristics (see Table 2). In the second combined model, institutional characteristics are inserted before student characteristics (see Table 3). By adding each group of characteristics gradually, one can examine the robustness of certain variables and whether they maintain their significance across multiple iterations of the model. For example, if a community-level variable maintains a statistically significant large effect on victimization rates as more variables are added to the model, one could assume this variable is robust. On the contrary, if the statistical significance or effect sizes of variables change across different iterations of the model, these characteristics might be less reliable predictors of campus sexual assault. As researchers could include a wide array of community, student or institutional categories in an analysis aiming for this level of granularity, future studies should work to refine this collection of variables in order to balance goals of comprehensiveness and validity. 

When looking at statistically significant factors across various stages of the model, the “percent of population 18-24 years old” and “median household income” variables are statistically significant at each stage. The coefficient for the “percent of population 18-24 years old” variable decreases as the model grows, while the “median household income” variable remains stable. While both coefficients are small, they both suggest that more young people in the population and higher median household income in the community lead to an increased number of sexual offenses at a respective college or university. The “percent students undergraduate” and “percent students full-time” variables also demonstrate statistical significance across all stages. Both variables have similarly small, but positive, coefficients, suggesting that increased numbers of full-time and undergraduate student are similarly associated with increased rates of sexual offense. The only institutional characteristic that shows statistical significance is the “in-state tuition” variable, suggesting that increases in a college or university’s in-state tuition leads to a small increase in the campus sexual offense rate.

To examine change in any results, Table 3 shows the model that adds institutional characteristics before student characteristics. Similar to the previous models, the “percent of the population 18-24 years old” and “median household income” variables maintained their effect size and significance. Within included institutional characteristics, “in-state tuition” also remained statistically significant with a small coefficient at both stages. However, the “highest degree awarded” and “public school” indicator variables demonstrated significance in the second stage, with community and institutional characteristics, but not in the full model. These results suggest that, without controlling for student characteristics, schools with a doctoral degree as the “highest degree awarded” reported 23 less officially reported sexual offenses per 100,00 students on average than other schools; more, they suggest that public schools report approximately 14 more officially reported sexual offenses per 100,000 students on average than private schools. Broadly, some of these variables have an obvious association with increased rates of sexual assault on campus (e.g. “percent students undergraduate”), but the analytical framework and early analysis shown above represent a basic model that future research should continue to build upon and refine in order to achieve maximum clarity on the problem and the factors that drive it.

Analytical Implications

There are important takeaways to this study, given the importance of reducing campus sexual assault. When examining community-level variables, the model in this analysis predicts higher rates of sexual assault at schools located in cities or counties with higher proportions of 18-24 year olds. While this seems obvious given the demographics of those involved in campus sexual assault, this still suggests that the problem is consistently prevalent amongst young people and might only increase in magnitude unless the cultural factors that perpetuate these behaviors are addressed. Given the digital nativity of the millennial generation, involved actors at the community and institutional levels could more easily reach out to young people through continued advocacy and awareness about campus sexual assault on social media and the Internet. Administrators at colleges or universities located in areas with particularly higher populations of young people could work with local policymakers, law enforcement officials and sexual victim advocates to determine ways that the community could better utilize technology to reach out to young people, educate them on what constitutes sexual assault and how to handle an incident if it occurs, and inform them of relevant prevention tactics. Trainings, such as bystander intervention training, could also be implemented to encourage individuals to recognize warning signs and intervene before a sexual assault occurs, an effort that simultaneously improves knowledge of prevention and increases the number of ‘capable guardians’ present on campus.

Innovative technology, and young people’s persistent engagement with it, also represents an important strategy to addressing student-level variables. Results from this analytical framework indicate that higher rates of sexual offenses occurred at institutions with higher percentages of undergraduate and full-time students. While this information, again, seems obvious given the demographics most likely involved in campus sexual assault, one could more reliably infer that students most affected by the problem are those that are younger and more frequently on campus. Institutional stakeholders, such as school administrators, Title IX officials, student organizations, and advocacy groups might then hold more events on campus, hand out pamphlets, or engage with on-campus student groups to help educate other students about campus sexual assault. School officials might also require mandated trainings for incoming or current students to inform them of sexual assault prevalence on campuses, prevention strategies, appropriate behavior and on-campus resources. This type of outreach, guided by data that suggests where and to whom to target efforts, could also be employed online – particularly, increased use of the school’s website, social media sites, smart phone applications, online surveys, and mass emails could help some students feel more comfortable in obtaining information and communicating through these mediums.

Within the institutional-level variables, there are fewer potential implications as only one such variable, “in-state tuition”, demonstrated consistent significance across the models. Only at one stage of the model were the “public school” and “doctoral degree” categories significant with large impacts. While these variables should be studied further before any conclusions are made, findings point to public schools and schools with higher degrees available as those with potential impacts on sexual assault rates. What about these public schools leads to higher rates of sexual assault on campus? Public institutions may have reporting structures that encourage more official reporting by victims, leading to a better gauge of the actual victimization rates on campus. On the other hand, private schools may just have less sexual victimization on their campuses due to smaller and more secured campus populations. More, if schools with higher degrees awarded show reduced rates of sexual assault, do doctoral or graduate students serve some role as ‘capable guardians’? If this relationship exists, schools could work to mobilize more ‘capable guardians’ on campus, such as increased campus law enforcement officers, faculty members, older students or students trained in intervention and prevention tactics. However, these represent only a sample of the possible questions future research could explore using the framework provided in this paper. While data from this analysis would likely change in future studies due to better data and a more refined model, this paper openly acknowledges the elementary nature of the analysis and its more purposeful attempt to construct a basic framework that questions what comprehensive factors might be at play when examining the drivers behind campus sexual assault.

Policy Implications: Looking Ahead

Increased Transparency and Quality of Data

This analysis points most strongly to the need for better data on sexual victimization rates and the contextual factors that influence them. Without more accurate institutional-level data, researchers will either have to use notoriously weak Clery data, which relies on institutions for official numbers, or compile different data sources themselves. Due to the inherent weakness and potential unreliability in the Clery data, results of this specific analysis could be indicative of actual differences in sexual victimization rates, or differences in the accuracy of reporting, reporting systems (i.e. systems that are victim-centered versus based on reports done by officials), or some combination of these factors. However, more comprehensive and reliable data may be coming: the 2014 White House Task Force to Protect Students from Sexual Assault highlighted the need for schools to conduct campus climate surveys to gather information from students and faculty rather than merely rely on official reports. If policymakers enact national standards to reduce campus sexual assault, such as mandating surveys or information-collecting processes annually, this contextual information could help researchers, practitioners and policymakers better understand the scope of the problem and set the stage for future research that could more accurately identify contributing factors. This knowledge would also allow policymakers and university administrators to create better policies to mitigate campus sexual assault and employ prevention strategies. The transparency of this information would also make it more difficult for administrators to cover-up or obscure higher rates of campus sexual assault than they are able to through officially reported avenues.

Increased Efficiency of Security and Prevention Efforts

If future analyses find a correlation between certain factors (community, student or institutional) and campus sexual assault rates, these findings could also impact broader policy solutions. Individual colleges employ different approaches to prevent campus crime; some colleges or universities use fully-accredited campus police forces (i.e. Virginia Commonwealth University), others rely on private security (i.e. Louisiana College), and some employ security officials that could be best described as ‘night watchmen’ (i.e. Dabney S. Lancaster Community College). If future studies could better pinpoint what environmental factors contribute to campus sexual assault (e.g. size of school, faculty gender ratio, location, community crime rate), such insight could inform decisions on what type of crime prevention strategies or security mechanisms might be best for an individual school.  

Implementation of National Standards

While school administrators should use this information to tailor prevention and security strategies to fit the unique needs of their institution, they should not be exempt from a broader responsibility to protect students from sexual violence. The variability in data across institutions illustrates the need for national standards of prevention and reporting; these expectations, such as mandating annual campus climate surveys, would ensure that schools are actively addressing the issue and working to reduce incidents of sexual assault, assist victims and protect students from future risk.

Need for Innovative Strategies and Solutions

Campus sexual assault is not a new issue, as studies have long found similar victimization rates among college women (i.e. Kanin 1967, 1970, 1977; “Not Alone”, 2014). However, more accurate and comprehensive data, and the increased prevalence of the issue amongst colleges in the millennial era, both suggest that viewing this issue through a generational lens may offer new solutions. For example, the effects of rapid technological advancements in the 21st century, epitomized by increasing use and availability of the Internet and social media, has connected people in unprecedented ways (Vito, 2013). Yet, this increased interconnectedness could facilitate in the spread of information that debilitates sexual assault prevention efforts on college campuses, such as cultural messages or conversations that exploit women and question what constitutes ‘real rape’. Mixed understanding or interpretation of what constitutes rape or sexual assault has contributed to the underreporting of sexual assault incidences, as many victims do not believe they were assaulted or feel pressured not to report the incident (Sloan & Fisher, 2010).

However, new-age technology and its seamless spread of information could also be used to open up new doors of communication, such as creating victim-centered, trauma-informed reporting channels.[vi] Many victims do not report due to distrust in officials (i.e. administrators, law enforcement), a problem that could be mitigated if technological innovations could refine reporting mechanisms to safely incorporate more information from victims. Joint efforts by the tech industry and advocacy groups – specifically by working with victim advocates, law enforcement, university administrators and other institutional stakeholders to leverage the benefits of technology in how it facilitates communication and information-collection - could create innovative tools for reporting, increase reliability and quality of data and better assist victims of campus sexual assault by providing new outlets for help. Groups could work together to produce a smartphone app where victims could report an incident and then quickly be provided with vital information, such as immediate steps to take to assure safety, services available to them on their specific campus or in the general area, and official reporting options with contact information. Given the traumatic nature of a sexual assault incident, particularly for young adults, developing new-age tools to provide quick and clear access to this information could be a first step in improving rates of underreporting and to gain a more accurate picture of how much sexual assault occurs across campuses.

Increased global connectivity is another characteristic of the millennial era that offers potentially untapped benefits. Policymakers, administrators and practitioners could look to research on campus sexual assault rates in other comparable countries to see if findings are relevant in the U.S. and learn from successful mitigation strategies. While less researched, other developed countries demonstrate a similar presence of campus sexual victimization. A 2014 survey of over 2,000 individuals from the U.K. found that 37 percent of female respondents and 12 percent of male respondents had faced unwelcome sexual advances while at U.K. universities (NUS Research, 2014). Given these similarly high prevalence rates, universities across different countries could work together in search of ways to mitigate sexual violence on their campuses. At the state and federal-level, policymakers could use this information to shape policies or fund pilot efforts that would produce better victimization data, make the reporting and adjudication process more victim-friendly, and determine the feasibility or success of national standards. At the institutional and local-level, administrators and practitioners could utilize information from international colleges to understand potentially influential factors, such as specific community or institutional characteristics, or improve policies and best practices, such as how to promote a campus culture that better mitigates sexual assaults.

Conclusion

While the prevalence of campus sexual assault and the subsequent attention paid to the issue has proliferated in the millennial era, many gaps in knowledge and data still exist. This study set the stage for future research to address an element of these gaps, the impact that multi-level characteristics of a college or university can have on its campus sexual assault rate. On the whole, this paper demonstrates the ability of future analyses, ideally equipped with better data and a refined model, to inform policymakers about what types of colleges and universities have higher sexual victimization rates, what exactly drives increased prevalence and what factors should be addressed when configuring solutions or use of resources. Schools could use findings from such comprehensive analyses to identify and address specific factors that are found to strongly influence rates of sexual assault, as well as enact more effective policies, trainings and best practices. Administrators could tailor outreach, prevention efforts, reporting mechanisms and security based on what characteristics or contextual factors lead to increased incidence.  Government agencies could also use this information to inform institution-level standards and requirements or assist institutions in being more proactive in how they respond to the issue.  Ultimately, these findings encourage future studies to question not only what drives sexual assault on college campuses, but what collection of characteristics influence sexual assault rates and why they do.

Despite the methodological weaknesses from which this analysis suffers, its findings still indicate that characteristics of the community in which a college is located, the student body that comprises it, and the institutional environment created by the school all may indicate increased or decreased risk of assault or the rate at which assaults are reported at that institution. The comprehensive set of characteristics and simple analytical model employed in this study considers both the individuals involved in the incident and the sociocultural factors that engender it, creating a foundation that can be easily built upon by future researchers. Policy implications arising from continued research of this nature speak to how stakeholders can utilize the benefits of the millennial era, in its increased connection and innovation, to determine post-millennial solutions, where strategies rely on more creative tools and more informed insight to reduce rates of sexual victimization on campuses. If actively pursued by policymakers and administrators alike, the unique features of 21st century technology and innovation could improve the quality of data collection, reporting systems, outreach and education efforts and support services for victims, all of which would lead to a stronger body of knowledge with which to move forward. Thus, the comprehensive nature of the analysis attempted in this paper, and the wide-ranging implications that accompany it, have the ability to not only create stronger responses to sexual violence on college campuses but also create a stable pathway toward reducing it.

Steven Tucker Keener is a third year doctoral student in the Public Policy and Administration Program at Virginia Commonwealth University. Steven earned a B.A. in Political Science, with a minor in Philosophy and Religious Studies from Christopher Newport University and a M.S. in Criminal Justice from VCU. Steven works as a research analyst for the Department of Criminal Justice Services, and has also worked as a research assistant for Dr. Charol Shakeshaft in the VCU School of Education. His research interests include campus and school safety policies, collegiate sexual violence, educator sexual misconduct in K-12 institutions and emergency preparedness on college campuses.

Contact Information: Steven Keener, Public Policy and Administration, L. Douglas Wilder School of Government and Public Affairs, Virginia Commonwealth University, 923 W. Franklin St., Richmond, VA, USA. E: keenerst@vcu.edu T: (540) 958-1023.