Introduction

In recent years, sexual violence on college campuses has become a pivotal issue. To illustrate, research has demonstrated that approximately 20 % to 25 % of women will be victims of attempted or completed sexual assault during their college years (Fisher, Cullen & Turner, 2000; McCaskill, 2014; Sinozich & Langton, 2014). Although more general concern with crime on college and university campuses has existed since the 1980s (Sloan & Fisher, 2010a, b), the greatest amount of attention from media, policymakers, and academics has been on sexual violence against college women (Sloan & Fisher, 2014). The impetus for this interest in violence against college women is not attributable to any singular event but instead to a steady stream of media coverage of date rape and sexual assaults occurring on campus (e.g., Stoeffel, 2014), combined with various groups’ efforts to socially construct campus crime – including violence against college women – as a new social problem (Sloan & Fisher, 2010a, b).

In an effort to address crime and violence on college campuses, Congress passed the Jeanne Clery Disclosure of Campus Security Policy and Campus Crime Statistics Act (20 USC §1092f) in 1990 (hereafter, Clery Act). Congress has amended the Clery Act several times since its initial passage, most recently in 2013 when President Barack Obama signed into law The Campus Sexual Violence Elimination (SaVE) Act, as part of the Violence Against Women Reauthorization Act (42 USC §13,701). The SaVE Act enhances existing aspects of the Clery Act by: 1) increasing transparency in the reporting of on-campus sexual violence, 2) guaranteeing enhanced rights for victims of violence who pursue disciplinary action against offenders, 3) setting standards for campus disciplinary proceedings, and 4) requiring institutions to provide campus-wide prevention and education programs (Kingkade, 2014). By October 2014, all Title IVFootnote 1 eligible postsecondary institutions were to “reflect a good faith effort for compliance,”Footnote 2 which is being enforced by the United States Department of Education under the requirements of the Clery Act.

For years, questions have been raised over institutional compliance with the Clery Act. As an unfunded mandate imposed on institutions, effected schools must devote significant resources to comply with the law. According to Richardson (2014), “College officials have continually argued that the [Clery Act] is a financial burden requiring additional paperwork and resources to comply with requirements that are complex and ambiguous” (p.4). (See also Gregory & Janosik, 2003; Lipka, 2009). With the additional requirements of the SaVE Act added to the mix, compliance with Clery may be even more problematic. To examine this issue, the present study explored whether, and to what extent, postsecondary institutions were in compliance with the SaVE Act during the 2013–2014 academic year.

Past Research

University campuses have traditionally been regarded as safe places that are largely free from crime. More recently, however, the pristine image of the “ivory tower” has been tarnished, and crime on university campuses has emerged as a serious issue (Armstrong, Hamilton & Sweeney, 2006; Fisher, 1995; Fisher, Daigle & Cullen, 2009, 2010; Nobles, Fox, Khey & Lizotte, 2013; Sloan, 1994; Sloan & Fisher, 2010a, b). Contributing to this perception is a large body of research showing that unacceptably high levels of violence – both sexual and nonsexual – is perpetrated against college women on campus, and much of this violence remains unreported to the authorities (Armstrong et al., 2006; Breitenbecher, 2001; Fabiano, Perkins, Berkowitz, Linkenbach & Stark, 2003; Fisher, 1995; Sloan, Fisher & Cullen, 1997).

The Clery Act was originally designed as “consumer friendly” legislation to address crime on college campuses by requiring postsecondary schools to make their crime statistics publically available (among other requirements). However, as a result of multiple amendments, Clery’s focus has shifted to requiring institutions to formulate policies aimed at preventing violence and protecting victims. Fisher, Hartman, Cullen & Turner (2002) argued the goals of the Clery Act were (and are), at least in theory, substantive in nature and include requiring universities to disclose their statistics, services, and programs to current and prospective employees and students. Yet, considering the federal government has provided an unfunded mandate to universities (expect for the possibility of obtaining specialized federal grants), the Clery Act, in reality, is more symbolic than substantive policy (Burke & Sloan, 2013; Fisher et al., 2002). The main purpose of the law seems to be placating interest groups that have argued university campuses are dangerous places that warrant administrative, legislative, and judicial intervention (Sloan & Fisher, 2010).

Clery Act

In its current form, the Clery Act requires all Title IV eligible, postsecondary institutions to comply with seven major disclosure requirements relating to crime and security on college campuses. These include: (1) disseminating an annual security report containing three years worth of selected crime statistics for the campus, policies and procedures for protecting the rights of victims of sexual assault, and overall security policies for the campus; (2) maintaining a public crime log documenting the “nature, date, time, and general location of each crime” and its disposition, if known; (3) disclosing crime statistics (gathered from campus police/security, local police, and/or school officials who have significant responsibility for student and campus activities) for incidents occurring on campus, in unobstructed public areas immediately adjacent to or running through campus, and at certain non-campus facilities including Greek housing and remote classrooms; (4) issuing timely warnings about Clery Act crimes that pose a serious or ongoing threat to students and employees; (5) devising an emergency response, notification and testing policy; (6) compiling and report fire data to the federal government and publish an annual fire safety report; and (7) enacting policies and procedures to handle reports of missing students.

Complying with Clery

The argument that the Clery Act is symbolic policy is buttressed when considering the penalties available for noncompliance and the low probability of sanctions that can actually be imposed on schools for failing to comply with Clery’s mandates. In fact, over the past 12 years, complaints alleging over 400 Title IV universities failed to comply with Clery resulted in only 62 audits of those schools’ Clery filings.Footnote 3 For those schools that have been audited and found non-compliant, fines are limited by Clery to $35,000 for every violation. In an effort to determine whether universities were discouraging reports of sexual assault or simply trying to hide them, Yung (2015) assessed the frequency of reports for sexual assault at four-year institutions that had been audited for non-compliance with Clery. Her results demonstrated that in the year the audit was performed, sexual assault reports increased by 44 %, and then decreased the following year, while reports for other crimes (robbery, assault, etc.) did not increase significantly during the same period.

Since 2012, there have been substantial efforts by policymakers to combat violence against college women. In 2014, President Obama launched the “It’s on Us” campaign in an effort to encourage personal responsibility and bystander intervention to aid in preventing and combatting sexual assault on campus. In response to the report presented by Senator Claire McCaskill (2014) to the Senate on campus sexual assault, a bipartisan group of senators proposed the Campus Accountability and Safety Act bill which, among many other changes, increased civil penalties institutions for Clery violations from $35,000 per violation to $150,000.

Clery Act Amendment : Campus Sexual Assault and Violence Elimination

According to Senator McCaskill (2014), highly-publicized campus sexual assault cases have only begun to shed light on the problem. In a report prepared by the United States Senate Subcommittee on Financial and Contracting Oversight, a survey of 440 four-year institutions found that 40 % of the universities failed to conduct an investigation on any reported sexual assaults in the last five years. Furthermore, many schools did not provide any sort of on-campus services for sexual assault victims and did little to address a culture that encouraged non-reporting of sexual violence.

In March 2013, President Obama signed into law an amendment to the Clery Act, known as the Sexual Assault Violence Elimination Act (SaVE) as part of the Violence against Women Reauthorization Act (VAWA). Originally independent of VAWA, the SaVE Act failed to pass, as many legislators were concerned that the proposed legislation stripped people who are accused of sexual assault of their constitutional rights while also failing to effectively remedy shortcomings in the Clery Act. In an effort to ensure eventual passage, Congress chose to include the SaVE Act within reauthorization of VAWA. Furthermore, the requirements of the act coincided substantially with recent Title IX requirements developed by the Department of Education for postsecondary institutions (Carter & Kirkland, 2013; Marshall, 2014).

The SaVe Act amendment revised Clery to go beyond the approach that mere awareness of past crime prevents future crime. The SaVE Act requires schools to provide more information on violence against women within their annual security reports, specifically, information on sexual assault, dating violence, domestic violence, and stalking. To be compliant with SaVE Act requirements, institutions must provide specific definitions of and statistics on, sexual assault, dating violence, domestic violence, and stalking. While Clery requires that institutions provide written rights and procedures for victims, the SaVE Act provides specific procedures for specific crimes, as well as requires schools to enhance and guarantee certain rights and protections to crime victims. Colleges and universities must also provide investigative procedures and guarantee enhanced rights of victims who are employees and students. Additionally, schools are required to provide information on bystander intervention programs that are available either at their school or elsewhere. Lastly, it requires schools to provide prevention programs that address sexual assault, dating violence, domestic violence, and stalking. While schools are not required to provide a particular form of prevention program, schools are required to have in place at least some type of program that addresses these crimes. Thus, similar to Clery Act requirements, schools are left with the discretion to choose specific forms of prevention and awareness programs (Carter & Kirkland, 2013; Marshall, 2014). It appears that the purpose of the SaVE Act was to go beyond simple symbolism of increasing perceived safety, and implement actual methods of prevention, risk reduction, and awareness.

Online Resources and Campus Crime

Research has demonstrated that college students use the internet more than any other source to find everyday information (Head & Eisenberg, 2011), particularly for sexual assault and health-related information (Escoffery, Miner, Adame, Butler, McCormick & Mendell, 2005; R. M. Hayes-Smith & Levett, 2010). Dameron, DeTardo-Bora & Bora (2009) assessed the availability and accessibility of universities’ campus safety website and Clery information. Of the 323 schools within the sample, they found that only 72 % had a public safety website, and of that 72 %, only 56 % made reference to Clery on their public safety website. To measure accessibility, Dameron and colleagues recorded the number of mouse clicks needed for one to start at each universities’ main page and end at the public safety website. On average, they noted two “clicks.” The researchers concluded that information on campus safety should be easier to access, and that there should be more uniformity in how information is reported.

In their study on the availability of women’s resources, R. Hayes-Smith & Hayes-Smith (2009) conducted a content analysis of sixty university websites. Specifically, they examined whether universities provided women’s centers, what resources were available, as well as the quality of those resources provided. The results of their study revealed that two-thirds of the universities did not have a women’s center and, while the majority provided some sort of resources on sexual assault and other issues, only a third of the universities’ resources were ranked as high quality. Furthermore, Krivoshey, Adkins, Hayes, Nemeth & Klein (2013), in their study of sexual assault reporting found that one-third of 105 Ohio universities provided no sexual assault policy, and only .03 % of those institutions provided a comprehensive definition of sexual assault that included all six offenses.

Most recently, Lund & Thomas (2015) conducted a content analysis of 102 universities to examine what information was available on sexual assault. To conduct a comprehensive review, the researchers examined all websites that were under the universities’ domain and assessed the availability and location of resources. The results showed that only 61 of the 102 schools provided a definition of sexual assault. Further, they found that over one-third of the institutions provided other information on sexual assault; however, Lund and Thomas also observed some emphasis of victim responsibility, coupled with a lacking emphasis on dispelling rape mythology.

Prevention, Risk Reduction, and Awareness

Fisher (1995) noted universities have a duty (based on either the tort of negligence or contract theory) to provide a safe campus that protects students. Furthermore, universities have a vested interest to protect both its employees and students, as crime on campus may affect its appeal to both potential students (Fisher & Nasar, 1992) and faculty members (Gover, Tomsich, Jennings & Higgins, 2011; Tomsich, Gover & Jennings, 2011). One of the most common types of crime prevention programs on university campuses has been to provide additional or more stringent guardianship. This can be accomplished through a myriad of methods, including the design of campuses (Cozens, Saville & Hillier, 2005; Merry, 1981; Taylor & Gottfredson, 1986), specifically, increasing lighting, landscaping, parking lot design, as well as emergency telephones and blue-light alarms. Moreover, guardianship can also be increased through nighttime escort services, campus patrols, as well as security in dormitories and other buildings used during the evening. Additioanlly, universities can provide sexual assault education that focuses on risk reduction and increasing awareness of the prevalence of sexual assault, particularly by acquaintances (Breitenbecher, 2001; Gidycz, Rich, Orchowski & King, 2006; Levine-MacCombie & Koss, 1986; Lonsway, 1996; Lonsway & Kothari, 2000). Furthermore, programs should also focus on self-defense skills (Gidycz et al., 2006).

Programs that are focused on risk reduction are often framed by lifestyle/routine activities; thus, providing methods for potential victims to change their behaviors. These are often associated with alcohol use and behavior, considering that past research has shown the two to be significantly correlated (Abbey, 2002; Benson, Gohm & Gross, 2007; Krebs, Lindquist, Warner, Fisher & Martin, 2009). While these programs are ostensibly beneficial, they provide no means of reducing potential offenders’ harmful behaviors. Furthermore, risk-reduction programs are often targeted towards women and can insinuate that women have the ability to prevent their own sexual violence victimization. These programs can also convey a message that women should be fearful of potential victimization. This coincides with past literature that has demonstrated women have a significantly higher fear of crime (for both personal and property crime) than men (Fisher & May, 2009; Gover et al., 2011; Tomsich et al., 2011). As a result, female students may become so fearful that their precautionary behaviors hinder their quality of life (McCreedy & Dennis, 1996; Tomsich et al., 2011; Wilcox, Jordan & Pritchard, 2007).

In addition to the aforementioned issues, there has also been concern regarding the presence of a rape-supportive culture, which is often evident within universities’ atmospheres (Armstrong et al., 2006; Burnett, Mattern, Herakova, Kahl Jr, Tobola & Bornsen, 2009). Such a culture can be indirectly affected by gender segregation, which occurs within many school activities, including Greek and athletic organizations (Franklin, Bouffard & Pratt, 2012; Schwartz & DeKeseredy, 1997; Schwartz, DeKeseredy, Tait & Alvi, 2001), as well as organizations that are predominantly male and promote masculinity, like ROTC (Scarce, 1997; Silva, 2008). There have been a number of proposed and implemented methods to dispel this culture, one being bystander education and awareness, which emphasizes men (and women) to step in and prevent or stop sexual violence (Gidycz et al., 2011). Additionally, other researchers have purported the importance of programs that are predominantly targeted at males that focus on methods of recognizing and avoiding rape-supportive behaviors (Breitenbecher, 2001; Fabiano et al., 2003; Foubert, 2000; Lonsway & Kothari, 2000), such as the perpetuation of rape myths (Burnett et al., 2009), which are defined as “prejudicial, stereotyped, or false beliefs about rape, rape victims and rapists—in creating a climate hostile to rape victims” (Burt, 1980, p. 217). Recommendations have been made to implement programs and policies focused on dispelling rape mythology (Breitenbecher, 2001; Fabiano et al., 2003; Foubert, 2000; R. M. Hayes-Smith & Levett, 2010; Lonsway & Kothari, 2000; Lund & Thomas, 2015).

While a myriad of programs have been proposed and/or implemented to address these various issues, concern regarding their efficacy has remained. To date, few programs have been effectively evaluated (Morrison, Hardison, Mathew & O’Neil, 2004). Of those programs that have evaluated, there have been mixed results. In an evaluation of a sexual-assault self-defense program, Gidycz et al. (2006) found that, compared to controls, those who participated in the program had higher levels of protective behaviors; however, there was no significant difference observed for rates of future sexual victimization. After the researchers revised the program to focus more on addressing barriers to initiate self-protective behaviors, the researchers found a significant reduction in sexual assault for program participants (Orchowski, Gidycz & Raffle, 2008). In another study, Gidycz et al. (2011) examined the effects of bystander intervention on sexually aggressive behaviors, and found that those who participated in the program reported less sexually aggressive behavior than the controls. In his evaluation of a rape-prevention program targeted toward fraternity men, Foubert (2000) noted significant declines for participants in both the likelihood of committing a rape and the acceptance of rape myths. In a recent study, Senn, Eliasziw, Barata, Thurston, Newby-Clark, Radtke & Hobden (2015) assessed the Sexual Assault Resistance Education program at three Canadian universities, which focused on methods of risk reduction through increased awareness, recognition of risky situations and behaviors, and methods of self-defense. The researchers noted a significant reduction in sexual assaults for first-year female students when compared to controls. Lonsway & Kothari (2000), in their evaluation of an educational program on acquaintance rape, noted that although there appeared to be significant changes in attitudes towards sexual assault initially, these effects were short lived. Thus, Lonsway and Kothari argued for the need for ongoing education and programs.

Current Study

To determine whether colleges and universities were in compliance with the SaVE Act in 2014,Footnote 4 we conducted a comprehensive examination of universities’ websites during the Spring of 2015. Our research was guided by five questions: (1) Are universities adhering to the guidelines of the Campus SaVE Act? (2) How accessible are universities’ online resources for sexual violence? (3) What types of programs do institutions most frequently offer on sexual violence? (4) Are student demographics, institutional characteristics, resources, and organizations associated with the availability of programs on sexual violence? (5) Are student demographics, institutional characteristics, resources, and organizations associated with compliance?

Methodology

To select institutions, we used the Department of Education’s Campus Safety and Security Data Analysis Cutting Tool.Footnote 5 With this tool, we selected only those institutions in the United States that were four year, main campuses, Title IV eligible, public and private, nonprofit (N = 2093). To ensure the sample was representative, we selected a 20 % sample (n = 444) of these schools that stratified by region (Northeast, Midwest, South, and West) and student population (less than 1000 students; 1000–4999 students; 5000–9999 students; 10,000–19,999 students; and 20,000 or more students). Despite our initial efforts to omit inapplicable schools, we omitted nine schools from the sample, including four that were solely online and five due to inoperable websites. This resulted in a final sample of 435 universities and colleges.

Procedure

The first step undertaken was to assess whether schools’ websites provided the basic information required by the SaVE Act in the annual security report required by Clery. To determine whether the selected universities and colleges were in compliance with the SaVE Act, we created a paper-and-pencil checklist that contained the requirements of SaVE (definitions, statistics, programs, and procedures) to assist in our review of website information. We also used the checklist to ensure verification of the information found at the school websites by creating an audit trail. To ensure clarity and accuracy of the measures, we conducted a preliminary analysis of 100 universities and colleges. Two of the researchers then met and discussed the results of the preliminary analysis, which was then used to specify the measures and better define compliance. The results of the preliminary analysis was then omitted, and the full sample was drawn (n = 435). One of the researchers then completed the checklists for all schools. After the completion of every 100 schools, the researchers would meet for purposes of peer debriefing. During these meetings, any items that were questionable for compliance would then be discussed to determine whether these met the criteria for compliance. Additionally, to ensure no errors existed, 10 % of the schools were reexamined to ensure there were no discrepancies, of which none were found. (If any discrepancies had been recorded, the full sample and procedure would have been reviewed and verified).

To examine the presence of information regarding the criteria of the SaVE Act, the first step was to search for a selected university using the Google search engine. Once a school’s main page was located, the researcher would then search the main website for the link to the university police or campus security. Once found, the researcher would then look for the security report. If the security report was located, s/he would then identify the presence of the required statistics, definitions, procedures, rights of faculty and students, and prevention program information. If the security report mentioned particular programs but then referenced an outside report or other information, the content of the information was also assessed to determine compliance with the SaVE Act. If we were unable to retrieve the information sought using this process, other search terms were then used, including Clery, Clery Report, Annual Security Report, Fire Report, Safety Report, Security Report, SaVE Act, Sexual Assault Violence Elimination Act, Crime, Sexual Assault, Rape, Stalking, Domestic, Violence, and Dating. The presence or absence of information was coded as “1= yes, information meeting this criterion was available through the university or college website” or “0 = no, no information meeting this criterion was available through the university or college website.”

With the exception of region, all data on institutional characteristics and student demographics were collected using the National Center for Education Statistics’ Integrated Postsecondary Education Data System (IPEDS).Footnote 6 We also used IPEDS to collect information on whether institutions had football and ROTC programs. For the remaining programs and policies, (whether institutions allowed alcohol on campus, had a women’s center, and had social fraternities and sororities), we first used the U.S. News & World Reports College Compass.Footnote 7 If this information was not available for an institution, we then used Peterson’s Guide to Colleges and Universities.Footnote 8 Lastly, if these data were unavailable for institutions through Petersons, we would use Google Advanced Search to search the institution’s website for key terms (alcohol: alcohol, alcohol policy, drugs; social fraternities and sororities: Greek, fraternity, sorority; women’s center: women’s center, center for women, women’s health, women’s facility). Neither U.S. News & World Reports nor Peterson’s provided information on Title IX officers; thus, we also used a Google advanced search to determine whether there was a Title IX officer at the school (using search terms Title ix, Title 9, Title nine).

Overall, we were able to find these data for each school, with the exception of six schools that each had more than 90 % missing data. These cases were excluded from the analysis, resulting in 429 schools. Additionally, there were missing data for financial aid (10 cases) and endowment (19 cases). After determining that these data were likely missing at random (those schools that were less likely to provide online resources on sexual violence were also less likely to report endowment and financial aid), we used linear regression imputation for the missing cases in these two variables.

Measures Footnote 9

Institutional Characteristics

Institutional characteristics included the institutions’ surrounding city/town population, region, campus housing, whether the school was public or private, tuition, endowment, and religious affiliation. For purposes of readability, within the regression models, population was reported in ten thousands, meaning each one-unit change represents a change of ten thousand people. Region had four categories: South, West, Midwest, and Northeast, which were each dummy coded. Campus housing was dichotomous (1 = yes, students can live on campus or 0 = no, students cannot live on campus). For public or private schools, we dummy coded the variable (1 = public and 0 = private). For tuition, we used in-state, fulltime tuition and fees for public schools and fulltime tuition and fees for private schools. For purposes of readability, tuition was reported in hundreds, meaning each one-unit change represents a change of 100 dollars. Similarly, endowment was reported by a hundred-thousand, meaning every one-unit change represents a change of 100,000 dollars. Lastly, religious affiliation was dummy coded with “1″ indicating the school was affiliated with a religion.

Student Demographics

We used enrollment data from 2014 to examine student population, age, sex, race, and receipt of financial aid. Population was measured as the percentage of students who were under 25 years of age. Sex was measured by the percentage of female students enrolled. Data on percentage of White, Hispanic, Asian, and Black students were collected, and then coded as the percentage of nonwhite students. We used the total percentage of students receiving aid as our measure of financial aid recipients.

Resources, Organizations, and Policies

As previously mentioned, past literature has demonstrated that organizations segregated by gender may indirectly encourage a rape-supportive culture on college campuses. Therefore, we aimed to examine compliance in context of institutions’ presence of these organizations (football, Greek life, and ROTC). All three were dummy coded measures with “1″ indicating the presence of each program.

Past literature has also examined the quality and availability of resources available within women’s centers on college campuses. Therefore, we included this as a possible factor associated with the availability of programs and overall SaVE compliance. This was also dummy coded with “1” indicating the presence of a women’s center.

Considering that the requirements of the SaVE Act coincide with Title IX requirements developed by the Department of Education for postsecondary institutions, we included the presence of a Title IX officer (dummy coded with “1” indicating the presence of an officer) as a possible factor associated with SaVE compliance.

Lastly, considering the relationship between alcohol and sexual assault, we included a measure of alcohol on campus, specifically, whether alcohol was allowed on campus for those over 21 years of age (dummy coded with “1” indicating that alcohol was allowed on campus).

Compliance

To determine overall compliance, we created a sum of the eighteen criteria for SaVE Act (α =  . 959). These criteria were: (1–8) definitions and crime statistics of sexual assault, domestic violence, dating violence and stalking; (9) procedures and policies regarding the disciplinary actions should one violate these procedures; (10) employees’ rights should they become a party to a crime; (11) students’ rights should they become a party to a crime; (12) the availability of one or more programs on sexual violence for new students; (13) the availability of one or more programs on sexual violence for new employees; (14–17) programs/workshops available for students that focus on reducing risk and/or increasing awareness of sexual violence; specifically on sexual assault, domestic violence, dating violence, and stalking; and (18) programs, workshops, and/or information on bystander intervention.

Accessibility

Accessibility of information was measured by counting the number of separations (or slashes) of resources’ websites from a university’s home page. The purpose of this was to provide an objective measure of accessibility to the information. Past research seeking to quantify accessibility to online information found at a particular website has used the number of mouse clicks required to reach the target information (Anney, 2013; Buhi, Daley, Fuhrmann & Smith, 2009; Dameron et al., 2009; Norman, Friedman, Norman & Stevenson, 2001). Arguably, slashes representing separations from a university’s homepage would provide a more accurate measure as this method eliminated concerns with testing effects. Meaning, that a person conducting an assessment of schools’ websites would gradually become more familiar with universities’ websites, and thus, know where to click, resulting in less “clicks.” Therefore, we used separations from a university’s home webpage to ensure that we had an objective measure of each school’s report from an institution’s main page. If the document (.pdf, .doc, .docx) did not follow standard html structure or correspond with the URL structure of the university website, we would then use the corresponding webpage in which we found the document and add one slash. (For instance, if ABC university’s Clery Report was found on the police website, yet the actual PDF’s web address was abc.sites/default/files, we would use the police website (www.abc.edu/student-services/police) and add one slash.

Analytic Technique

To determine whether institutions were adhering to the guidelines of the Campus SaVE Act (RQ1), we examined frequencies of each criterion for compliance, as well as the overall average count of criteria schools satisfied. To assess the accessibility of online resources at schools and universities (RQ2), we reported the mean number of slashes for programs, definitions, statistics, and the Clery Report. To determine what types of programs were frequently offered on sexual violence (RQ3) we reported those that were most commonly available at institutions. Additionally, we provided descriptions of some of the more prevalent and/or unique programs reported. We used binary logistic regression to examine the association of institutional characteristics, student demographics, and institution’s resources and organizations with the presence of programs/workshops on sexual assault, domestic/dating violence, and stalking (RQ4). For overall compliance, we determined that the dependent variable was an overdispersed count variable. After further examination of the distribution, we used a negative binomial regression to examine the association of institutional characteristics, student demographics, and institution’s resources and organizations with overall compliance (RQ5).

Results

Compliance, Institutional Characteristics and Student Demographics

Descriptive statistics (Table 1) show that on average, institutions within the sample met 9.68 of the twenty criteria. On average, surrounding cities had a population of over 262,000. Institutions were relatively equal across regions, with 28.4 % in the Northeast, 26.6 % in the Midwest, 24.0 % in the South, and 21.0 % in the West. The majority of schools provided campus housing (89.7 %) and over one-third were public universities (39.6 %) and were affiliated with a religion (36.4 %). Colleges’ and Universities’ mean tuition was $20,492. The majority of schools had Greek social organizations (53.6 %), allowed alcohol on campus (59.4 %), had ROTC programs (59.9 %), and had a designated Title IX Officer (92.3 %). Almost half of the schools had a football program (45.0 %) and almost a third had women’s centers (32.6 %). On average, students within each institution were predominantly under 25 (68.7 %), female (57.5 %), white (58.3 %), and receiving financial aid (72.4 %).

Table 1 Descriptive Statistics
Full size table

In addition to looking at overall compliance, we examined the distribution of each criterion (Table 2). The majority of schools provided definitions and statistics on sexual assault (74.4 % & 85.0 %), domestic violence (62.2 % & 66.1 %), dating violence (61.2 % & 65.4 %), and stalking (61.9 % & 65.1 %). Less than half of schools provided any reference to or information on sexual assault programs (49.8 %), and less than a third referenced programs on domestic violence (27.0 %), dating violence (28.1 %), or stalking (20.3 %). While the majority of schools provided disciplinary procedures (63.2 %), written rights to employees (65.3 %) and students (66.2 %), only a third provided information on bystander intervention (33.3 %), less than half provided prevention programs for new students (39.2 %) and less than a third provided prevention programs for new employees (29.2 %). Overall, 30.1 % of schools had low compliance, and only 11.2 % had high compliance.

Table 2 Compliance Criteria
Full size table

Accessibility

To determine the accessibility of information required under the SaVE Act, we examined the mean number of slashes within each web address where the information was located, typically within the annual security report. As shown in Table 3, there was little difference in the mean number of slashes in the URLs where the information was found. On average, the annual security report had 3.38 separations (slashes) from the main website. Generally, definitions and statistics were found in the annual security report; however, at times these were located in a linked addendum, thus adding another separation. As a result, the average for definitions, statistics, procedures, rights, and programs were slightly higher (ranging from 3.28–3.55). While the mean separations showed overall high accessibility, a few outlying schools had 6–9 separations from the main webpage (Table 3).

Table 3 Accessibility Descriptive Statistics
Full size table

Types of Programs Offered

Table 4 presents frequencies of the most common types of programs (with at least 9 (1.5 %) of institutions reporting that they provide the program). It is important to note that the numbers are not necessarily indicative of the number of schools that offer the program, simply that institutions referenced these programs online, and they were one of the first three referenced that we found for sexual assault, stalking, domestic violence, or dating violence.

Table 4 Most Common Programs Advertised
Full size table

The most commonly listed program was Rape Aggression Defense, with 11.7 % of the schools referencing this program within their online materials. The second most common was Haven, which is an online training on sexual violence. While other common programs addressed sexual violence though awareness (Take Back the Night and Clothesline Project), the “Think About It” program focuses on preventing sexual assault through the reduction of risky behavior.Footnote 10 Some universities would reference general sexual assault, domestic violence, or stalking awareness and prevention programs without giving much detail about the specifics of which crime was being addressed by the program and thus were classified as sexual assault prevention, domestic violence prevention, etc. In addition to the programs presented in Table 4, there were another 67 programs that at least two schools advertised as providing prevention of violence against women. Moreover, there were another 79 unique programs that only one school offered. This may show a lack of standardization on its face; however, many of the programs seemed to address similar topics (consent, acquaintance rape, date rape drugs, etc.)

To provide a more comprehensive illustration of some of these programs, Table 5 presents short descriptions of programs that were less explanatory by name. As one can observe, some of these programs that went beyond simple awareness and encouraged changes in behavior (for both potential victims and offenders) by incorporating self-defense, pledges to protect oneself and others, and other similar themes.

Table 5 Program Descriptions
Full size table

Factors Associated with Programs and Overall Compliance

Table 6 provide three logistic regression models that examine what factors were associated with institutions that provided online reference to one or more programs on sexual assault, domestic/dating violence, and stalking (Since many programs on dating and domestic violence overlapped, we chose to collapse these into one variable for this analysis). Both South and West were negatively associated with providing programs on sexual assault, dating/domestic violence, and stalking, meaning that institutions located in the South and West were less likely to provide online references to programs on sexual violence than schools in the Northeast. Student population was positively associated with programs on sexual violence, indicating that higher student populations were more likely to provide online references to programs they offered on sexual assault, domestic/dating violence, and stalking. The presence of women’s centers was also positively associated with dating/domestic violence programs, indicating that schools with women’s centers were more likely to make reference online to dating/domestic violence programs. Conversely, financial aid was negatively associated with dating/domestic violence programs, schools with a higher percentage of students receiving aid were less likely to reference online to a dating/domestic violence program. All other institutional and student variables lacked significant associations with sexual violence programs across all three models.

Table 6 Logistic Regression Models for Program Availability
Full size table

To examine overall compliance, we used a negative binomial model, as compliance was a count variable that was over dispersed. The results are presented in Table 7. Similar to programs, South was negatively associated with overall compliance and student population was positively associated with compliance. Additionally, ROTC programs was positively associated with overall compliance. All other institutional and student variables lacked significant associations with overall compliance.

Table 7 Negative Binomial Regression Model for Overall Compliance
Full size table

Discussion

The requirements of the Campus Sexual Violence Elimination Act (SaVE Act) appear to address some of the issues critics have identified as shortfalls with the Clery Act. Rather than requiring postsecondary institutions merely report data on campus crime, the SaVE Act stresses postsecondary institutions undertake efforts to prevent and reduce the risk of crime, especially violence against women on campus. Furthermore, the legislation requires that schools define what constitutes certain types of violence against women and specify the rights of students and employees, should they become crime victims.

Our first research question was whether institutions were adhering to the guidelines of the Campus SaVE Act. Overall, we observed a “good faith effort” to comply by approximately half of the universities. Specifically, we found that on average, schools were meeting less than 10 of the criteria. Overall, 35.9 % of all institutions had high levels of compliance, yet only 11 % had full compliance. While universities were required to reflect a good faith effort of compliance by October, 2014, we observed that 30 % of universities were completely noncompliant or met less than four of the eighteen compliance criteria. Some of this may be attributable to the time of data collection (spring 2015), which did not provide a substantial break between date of required compliance and data collection.

Our second research question focused on the accessibility of resources for students and employees. On average, resources were three to four separations from the university main page; however, there were outliers that had up to nine separations from the main page. While this provides a measure of accessibility, overall accessibility, particularly for students, may be hard to quantify. Considering that past research has shown the majority of students are not even aware of the report or the Clery Act (Janosik, 2001; Janosik & Gehring, 2003), then they likely would not realize that the report contains crime statistics, policies, and procedures—considering its title does not include the word crime. Moreover, even if a student is aware of these laws, they may not think about how to go about accessing such information if the student has just been a victim of a crime. Considering that on average, SaVE Act information had at least three separations from the universities’ main pages, a student may not be able to find this information easily in a heightened emotional state.

Our third research question was regarding what programs institutions most commonly referenced online. We observed that programs generally fell into educational/awareness programs and risk-reduction programs. Specifically, a number of institutions referenced domestic violence and sexual assault awareness (as either an awareness month/week or a general program to promote awareness). Similarly, both the Clothesline Project and Take Back the Night are programs whose purpose is to bring awareness to the prevalence and harm of sexual violence. Haven, a program run by the company Everfi, provides online training for students and employees that provides methods of risk reduction, and also educates students and employees about procedures and information at each campus. Other programs focused on methods of risk reduction and recognition of violence. The most popular risk reduction program was Rape Aggression Defense, which focuses on methods of self-defense, particularly for women. Other risk-reduction programs included “Think About It,” which focuses on risky behavior and risks of victimization, as well as recognition of sexual violence, and methods of supporting victims of sexual violence. Similarly, “healthy relationships” teaches students about recognizing the characteristics of healthy versus unhealthy relationships, as well as healthy decisionmaking for current and future relationships.

Within the most common programs, we saw no programs that focused solely on changing rape-supportive culture. While Haven does have some focus on bystander intervention, as does Healthy Relationships, all other common programs were primarily focused on methods of risk reduction for potential victims. While not within the most common programs, we did find that some universities offered programs that focused on giving and recognizing consent, such as “Can I Kiss You?” Further, four schools provided online information on “Men against Violence,” where men are brought together to rally against violence. What we did not observe, however, were programs that had a primary focus on dispelling rape myths. While these may be referenced within some programs, we did not find this to be a concentration within any programs. Thus, while previous research has recommended programs that focus on diminishment of a rape-supportive culture atmospheres (Armstrong et al., 2006; Burnett et al., 2009), and a shift of programmatic emphasis from women’s behaviors to men, such recommendations do not appear to have been implemented by universities.

Our fourth question was whether institutional characteristics, resources, and organizations and student demographics were associated with program availability and overall compliance. We first examined how these were related to providing information on programs, considering that these showed compliance, but also required more initiative and formal implementation than the others, which generally required the reporting of information and/or policies. We found that region and student population were significantly associated with online references to all three types of programs and overall compliance. Additionally, the presence of a women’s center was also positively associated with programs on dating and domestic violence, ROTC programs were positively associated with overall compliance.

It may be the case that schools with more students have more resources to ensure compliance with the SaVE Act than those with smaller student bodies. Furthermore, schools with larger student populations may feel more pressure to comply with both Clery and the SaVE Act because larger universities tend to have more well-known “brands” and thus occupy a larger place in the higher education landscape. Lastly, regardless of the crime rate, events of crime are likely to be higher at larger universities; thus, there would be a greater likelihood that larger universities would be spotlighted for criminal events, which would increase the likelihood of an audit. For instance, rates of sexual assault would be equal between a university with 1000 students and a university with 20,000, even though one sexual assault would have occurred at the small school, versus twenty at the large university. Being aware of the likelihood of numerous criminal events may motivate administrators at larger universities to be proactive with compliance with campus crime mandates. Regarding women’s centers association with dating/domestic violence programs, it may be that the presence of women’s centers may indicate a culture at universities that supports the safety and health of women. Additionally, one of the purposes of many women’s centers is to provide workshops on issues of safety and health; thus, these types of programs may have been in existence prior to the requirements of SaVE. However, it is also important to note that similar to Hayes-Smith & Hayes-Smith (2009), we found that only one-third of our sampled schools had a women’s center—demonstrating the plausible lack of development in women’s centers in the past seven years.

The positive association between ROTC programs and compliance may be attributable to the determining the motivations behind institutions’ having ROTC programs (particularly those who have consistently provided these program). Specifically, it may be an indication of support for federal government and/or patriotism; thus, adherence with polices set forth by the federal government would fall in line with that motivation. Conversely, it may be that schools providing a male dominated program may indirectly encourage a rape supportive culture. Thus, compliance with the SaVE Act may be a preemptive effort on the institution’s part to reduce the rape supportive culture, even if only symbolically.

The fact that other institutional characteristics were not significantly associated with compliance may demonstrate the most interesting findings. These findings indicate that those schools who have organizations that are indirectly related to a rape supportive culture (football and Greek life) and allow alcohol on campus are not even complying with the law at a significantly higher rate than those without these characteristics. Thus, institutions with, ostensibly, the highest risk of sexual assault on campus are doing no better than other institutions at complying with the law, let alone going beyond compliance in an effort to reduce the risk of sexual violence on their campuses. Moreover, it appears that other resources (women’s center and established designation of a Title IX officer) provides no increase in overall compliance for schools, regardless of risk.

The variability in compliance supports the argument that both Clery and SaVE are little more than “feel good” symbolic policy, as no federal funding has ever been provided to help schools implement Clery, let alone SaVE Act requirements. Regardless of the intentions of the SaVE Act, the biggest issue surrounding the legislation will arguably be postsecondary institutional compliance. Considering that disclosure of sexual assault statistics has been a requirement of Clery for over twenty years, one would likely assume that most, if not all, schools are providing that information. However, even though Clery has required universities to report sexual assault statistics, within our sample, we found that 15 % of institutions did not provide these statistics.

Limitations

This study is not without its limitations. First, as previously mentioned, there was not a significant lag between required compliance (October 1st, 2014) and data collection (Spring Semester, 2015). Thus, there may be increased compliance within later years. Second, while efforts were made to ensure accuracy of available information, findings are still subject to error. Third, while we reported up to twelve programs for each school, there are likely programs that were not recorded, due to a high frequency of programs at the same university. Further, the availability of programs was solely assessed by whether they were marketed online. There are likely other methods of advertising these programs (fliers, pamphlets, emails, announcements) that were not advertised online. However, for purposes of compliance, mention or reference of these programs should be provided within the annual security report. Considering these limitations, we encourage caution in interpretation of our results.

Recommendations

Considering these findings, we believe there are a number of steps that universities, as well the Department of Education (and other federal organizations) can take to increase compliance, as well as overall effectiveness of SaVE guidelines. First would be to increase the accessibility of information by providing a descriptive link on the universities’ main page to all information and resources focused on crime (programs, statistics, procedures, etc.). This recommendation coincides with Dameron et al.’s (2009), that a link to crime information be provided on every university webpage. Second, we recommend changes in how information on crime is marketed towards students. The Annual Security Report was designed with compliance in mind—not students, as it is an administrative document (Lund & Thomas, 2015). It is most often lengthy, with detailed explanations of how the institution is meeting Clery requirements. The report is marketed towards auditors - not students. Thus, we call for a summary report of important information for students that is marketed to students. This should provide the most pertinent information on policies, procedures, statistics, and programs, while also providing links to more detailed information. We recommend that this be designed similarly to recruitment products. As Hayes-Smith & Levett (2010) stated, resources are only effective if students read them; thus, they must be available and marketed with students in mind. Similarly, we also recommend that programs on sexual violence be available within one website—regardless of what organization is offering them.

Our fourth recommendation is to increase funding opportunities to for compliance and development of programs to reduce risk and prevent crime. While the National Institute of Justice, through the Office on Violence Against Women, has offered grants to develop and implement prevention programs on sexual violence, in 2013, only twenty-eight universities were funded.Footnote 11 Further, while increasing funding, we also recommend, in conjunction with Senator McCaskill’s bill, an increase in the fines for each violation of Clery and its amendments, as well as an increase in the frequency of audits. Additionally, as previously mentioned, we observed vast variability in the types of programs that were offered. Considering that there have been few rigorous evaluations of programs on sexual violence (Morrison et al., 2004; Senn et al., 2015), universities need to ensure that they are implementing programs that are effective for their student population. Thus, it may be fruitless to require all schools to implement the same programs. Determining effective programs can only result from continual evaluation of various programs at multiple universities. Similar to Lund & Thomas (2015), we believe there is a need for programs on sexual assault (and other types of sexual violence) to focus on the perpetrators—not the victims. Therefore, in addition to increased funding for general compliance, we puport that there should be additional funds granted to universities to assess programs’ effectiveness on attitudes and behaviors, particularly for programs that support the diminishment of a rape-supportive culture and rape mythology. Lastly, we would recommend that universities each create a women’s resource center that has a strong focus on sexual violence. If resources are available solely through public safety, universities may convey a message that institutions’ role in criminal acts is simply a reactive one. Providing a women’s center would allow universities to address sexual violence through proactive and reactive measures, as well as through counseling. This could start to convey a message to students that preventing sexual violence is not solely the victim’s responsibility, but instead the responsibility of the campus community. Within such a center, students would be able to address concerns and fears in an, ideally, safe environment (Burnett et al., 2009; R. Hayes-Smith & Hayes-Smith, 2009).

Conclusion

To combat campus crime and violence against women, it is important to approach the needs of all involved and consider all consequences of each action. Violence against women is a vast problem on college campuses that can only be combatted with heightened and consistent awareness, methods and programs of prevention, and support and advocacy for victims. Yet, this cannot be fully achieved until there is full compliance among all institutions.