Women with violent partners may seek a variety of services and protections to reduce their level of danger and exposure to further victimization. State laws provide criminal sanctions for violence when criminal complaints are filed by victims. However, many women seek protection through civil rather than criminal remedies. Civil actions can include divorce, legal separation, and protective orders, which are sometimes known as “restraining orders.” Protective orders were developed to provide partner violence victims with a way to prohibit contact, or at least violent contact, by their offending partners (Finn 1989). In general, women seeking protective orders report a history of severe violence (Carlson et al. 1999; Gondolf et al. 1994; Keilitz et al. 1997; Klein 1996; Ptacek 1999). However, beyond research showing that women who seek protective orders have a history of severe violence, there has been limited research about the characteristics of women who obtain protective orders and their views of the effects of protective orders. Increased information about women who obtain protective orders can inform the development of services and may indicate some ways the justice system could improve the implementation of protective orders.

One of the important questions in the literature is whether there are differences in partner violence between cultures and ethnicities. The literature suggests that rates of intimate partner violence are similar across various cultural groups such as women living in rural and urban areas (Rennison and Welchans 2000) and for White and African American women (Tjaden and Thoennes 2000). For example, estimates from the National Crime Victimization Survey (NCVS) suggested that 10 per 1,000 urban women experience partner violence in the prior year compared to about 8 per 1,000 rural women, and that this trend was similar from 1993–1998 (Rennison and Welchans 2000). Similarly, the National Violence Against Women (NVAW) survey found that approximately 8% of both White and African American women reported rape by a partner, about 24% reported physical assault by a partner, and about 5% reported stalking victimization by a partner in their lifetime (Tjaden and Thoennes 2000). Even though rates of intimate partner violence may be similar across women from different cultural contexts, the experience of intimate partner violence may be different (Kasturirangan et al. 2004). Culture can be defined as the sharing of similar beliefs, practices, values, norms, and behaviors which are used to initiate and maintain behavior among group members (Hammond 1978). In addition, cultures can be regionally, ethnically, and/or racially bounded. This study uses race and geographic area as proxies for cultural groups, resulting in three different cultural groups.

Specifically, this study offers descriptive information for rural White, urban White, and urban African American women on relationship characteristics, victimization experiences, protective order stipulations and violations, and self-reported effectiveness of the protective order. The primary purpose of this paper is to better understand the similarities and differences in experiences of partner violence among the three groups of women. Understanding similarities and differences among women who obtain protective orders against violent partners for various cultural groups may be important for agencies when meeting the needs of women from a variety of backgrounds.

Materials and Methods

Participants

The total study sample was 757 women (rural, n = 378, and urban, n = 379) who were interviewed between February, 2001 and November, 2003. For purposes of this analysis, participants were categorized into three groups based on area and race. Because only seven women in the rural area were of a race other than White, only one group was possible for the rural area (rural, White, n = 371). The numbers of participants who were of races other than White in the rural area are the following: Hispanic (n = 2), Asian (n = 1), Native American (n = 1) and bi-racial (n = 3). The seven non-White women were dropped from the analysis because of small group size. Of the urban subsample, 254 were White, 103 were African American, 7 were Hispanic, 2 were Asian, 4 were Native American, and 9 were bi-racial. Because of the small number of participants in the racial categories other than White and African American, these cases were also dropped from analysis. The final sample for the analysis consisted of 371 rural White women, 254 urban White women, and 103 urban African American women.

Procedure

Researchers obtained permission from four court jurisdictions (three rural and one urban) to recruit female petitioners for the study after they had been granted protective orders. To be eligible for the study, participants had to be female, 18 years old or emancipated, and have obtained a protective order (domestic violence order [DVO]) against a male intimate partner within six months of entering the study. The average length of time between the DVO being issued and entry into the study was 39.2 days (SD = 45.8). Of the women approached in court, 83.5% provided contact information, 2.4% initially refused participation, and 14.1% took information about the study but did not provide contact information. Of those with valid contact information which were actively pursued by the study staff, 70% completed the interview. Out of the 30% not interviewed, 7.7% decided not to participate and 22.3% were never successfully contacted and/or scheduled. Interviews lasted an average of 3.5 h, and began after women gave informed consent. Participants were compensated for their time and, if applicable, travel and child care costs. All information presented in this paper was self-reported.

The three rural counties had a rural population between 78 and 100% as classified by the 2000 census. The urban county had only 4% of its population defined as rural by the 2000 census. According to the 2000 Census (http://www.census.gov/) the total population for all three rural areas was 97,108 (ranging from 25,277 to 42,441) while the total population for the urban area was 260,512. African American women constituted 27% of the overall urban sample for this study, which represents a higher proportion of African Americans than is found in the overall African American population in the urban county (14%). The African American population across all three of the rural counties was 3% (U.S. Census 2000).

Psychological abuse and physical abuse were measured by a combination of several instruments including the Revised Conflict Tactics Scale (CTS2; Straus et al. 1996), from a study examining protective order petition incidents (Harrell et al. 1993), from Tolman’s Psychological Maltreatment of Women Inventory (PMWI) (Tolman 1989, 1999), as well as through extensive pilot work with the target population (Logan et al. 2003). Five psychological abuse subscales (verbal abuse, degradation, jealousy and control, symbolic violence, and serious threats) were created by grouping questions from the CTS2 and the PMWI, based on several studies of psychological abuse dimensions (Follingstad and DeHart 2000; Kasian and Painter 1992; Marshall 1992; Straus et al. 1996; Tolman 1989, 1999). A question about stalking was adapted from the National Violence Against Women Survey (Tjaden and Thoennes 1998). Two physical abuse subscales (moderate and severe physical violence), sexual insistence, sexual assault, and injury scales were based on questions from the CTS2 (Straus et al. 1996). Sexual insistence consists of partners insisting (without using physical force) on intercourse when women did not want to have sex. Sexual assault included threatened and forced sex. Participants’ responses to an open-ended question about the effectiveness of the protective order were categorized and independently coded by two researchers. Discrepancies were resolved until there was 100% agreement.

Analysis

Descriptive analyses were used to examine group characteristics. Post hoc analyses were conducted with Tukey (HSD) and z test of proportions. Because of the large number of comparisons, only findings that were statistically significant at p < 0.01 were noted.

Results

Demographics

As Table 1 shows, the mean age of the total sample was 31.6, with no difference between groups. No significant group difference was found for the number of children, or the number of children living with the participant. There was a linear trend for marital status, with more African American women reporting they had never been married than the other two groups, and more urban White women reporting they had never married compared to rural women. Compared to urban White women, fewer rural women reported having a high school education or General Equivalency Diploma (GED) and rural women had a lower average annual income. More rural women reported being unemployed than urban White or African American women. When the sources of income or support in the preceding year were examined, not surprisingly, fewer rural White women had received income from work compared to the urban groups. Significantly fewer urban White women received income from Temporary Assistance for Needy Families (TANF) compared to rural White and urban African American women. Two thirds of rural White women were on Medicaid, which was significantly higher than urban White women and urban African American women. Further, nearly seven out of ten rural White women received food stamps, which was significantly greater than the percentage of urban White and urban African American women. Fewer urban African American women had received Social Security Insurance (SSI) or Social Security Disability Insurance (SSDI) compared to rural White women. Income from a spouse or partner did not differ by group. Looking only at women who had minor children, a significantly smaller percentage of rural White women received child support when compared to urban White women.

Table 1 Demographics
Full size table

Relationship Characteristics

As Table 2 indicates, more rural White women had been married to their partners than women in either urban group, whereas more of the urban women (White and African American) had cohabitated with their partners. Further, more rural White women had minor children in common with the partners against whom they had obtained the protective order (DVO partner) than women in either urban group. The length of the relationship with the DVO partner was significantly longer for the rural White women when compared to the urban White women and urban African American women. Moreover, significantly more urban African American women had been involved with their partners for less than 1 year when compared to rural White women.

Table 2 Relationship characteristics
Full size table

When relationship length was examined more closely, the average amount of time that the couples dated was not significantly different by group (averaging just over 1 year), even though the actual amount of time living together was significantly different. Participants were also asked to estimate how long after the relationship began that they first experienced abuse from the DVO partner as well as the first time they felt fear of the DVO partner. There were no differences by group on either of these variables with women reporting first feeling fear about 2.4 years into the relationship and estimating the first incident of abuse occurring about 2 years into the relationship.

More urban women in both groups reported living with their partner at the time they filed for a protective order when compared to the rural women. The majority of all the women reported being separated from their partner at the time of the interview, however, significantly more of the rural women reported being separated at the time of the interview than both of the urban groups. Further, about the same percentage of the three groups reported being very financially dependent on the DVO partner at the time of the interview.

Table 3 includes information about women’s perceptions of their relationship with the violent partner or ex-partner. About half of all three groups indicated that, looking back, they were not at all satisfied with the relationship. However, a small percentage of all three groups indicated they were satisfied or extremely satisfied with the relationship. Although there were no significant differences by group for satisfaction, there were significant differences in how often women wished they had never become involved with that partner. More rural women indicated they frequently wished they had never become involved with that partner compared to both of the urban groups. In addition, overall, a large proportion of all three groups of women indicated they confided in their DVO partner frequently; however, more rural women indicated they rarely confided in their partner when compared to the two urban groups.

Table 3 Participants’ relationship satisfaction
Full size table

Victimization

Partner victimization in the past year by group is displayed in Table 4. High percentages of all three groups experienced all types of psychological abuse (verbal abuse, degradation, jealousy and control, symbolic violence, and serious threats) and physical violence (moderate and severe). Approximately 20% of all three groups reported sexual assault by the DVO partner. There were significant differences by group with more rural women reporting symbolic violence (e.g., threats to harm pets or others close to her, destruction of property, threatening to hit or throw something at her) than the urban African American women. More rural women also reported sexual insistence by their DVO partners than urban African American women.

Table 4 Partner victimization experiences in the past year of the relationship with the DVO partner
Full size table

When examining the number of different tactics within each subscale, rural women reported more tactics within the categories of degradation, jealousy and control, and serious threats than women in the two urban groups. Rural women also reported more verbal abuse and symbolic violence tactics than African American women.

Protective Order Stipulations, Violations, and Perceptions of Effectiveness

Protective order (i.e., Domestic Violence Order) stipulations are shown in Table 5. Two categories of protective orders are issued: (1) No-contact orders, which prohibit all forms of communication from the respondent to the petitioner, including phone calls, letters, and email; and (2) No-violent-contact orders which prohibit further acts of violence or threats of violence by the respondent toward the petitioner, but allow the respondent to contact the petitioner. Fewer rural White women had obtained no-violent contact orders compared to women in the two urban groups. Furthermore, significantly fewer urban African American women obtained a no-contact order than urban White women. Of those who were granted a no-contact protective order, footage restrictions were placed on the partners of 95.6% of rural White, 28.5% of urban White, and 26.7% of urban African American participants. Thus, the provision of no-contact orders and footage restrictions was more common in the rural counties than in the urban county. Stipulations for the respondent to vacate the premises were more common for the rural White women when compared to both groups of urban women.

Table 5 Protective order stipulations
Full size table

Stipulations related to child custody and support issues were examined for participants who had at least one minor child in common with the DVO partner. In these cases, results show that temporary child custody was awarded in the majority of cases for rural White women (60.0%), whereas significantly smaller percentages of cases among urban White women (25.4%) and urban African American women (17.9%) involved temporary child custody provisions (see Table 5). Of those cases that included stipulations about temporary child custody, custody was awarded to the petitioner (the participant) in the vast majority of cases in each group (94%). Temporary support was awarded to one-sixth or less of the women who were awarded custody of a minor child in common with the DVO partner, with no significant differences by group. More rural women reported their partners were ordered to attend counseling compared to both urban groups, and more urban White women reported their partners were ordered to counseling when compared to African American women. Significantly more rural women (over one-fourth) were ordered to attend counseling when compared to women in the urban groups.

Reports of protective order violations and perceptions of the protective order’s effectiveness are presented in Table 6. Overall, about a quarter of all three groups reported their partner or ex-partner had violated the protective order with no differences by group. Approximately 59% of the women reported the DVO partner had ever been in jail. A little over one out of five of the women overall reported that their partner had been in jail in the preceding 30 days. No differences in percentages of partners ever or recently in jail were found.

Table 6 Perceptions of the protective order violations and effectiveness
Full size table

The majority of women reported they felt fairly or extremely safe from the DVO partner, although more rural women reported feeling less safe compared to the other two groups. Consistent with the indication of feeling safe, many women reported they felt free from the DVO partner. However, between 20 and 30% of the groups reported not feeling free from the partner, with no significant between group differences. Overall, the majority of the women indicated they believed the protective order was fairly or extremely effective because: (1) The DVO partner had not bothered her since the issuance of the order (48.3%); (2) She believed that the DVO partner was afraid of the consequences of violating the order (31.2%); (3) Violence had decreased (but not stopped) (8.2%); (4) The DVO partner was incarcerated and could not bother her (6.7%); (5) The order had given her a sense of power or safety (6.1%); and/or (6) She thought that her DVO partner understood that she was serious about stopping the abuse (5%). There were no significant group differences in reasons for rating the protective order as effective. Among the 12.5% of the women who rated the order as not effective, the most common reasons cited for rating the protective order as not effective were the following, with no differences by group: (1) The petitioner violated the order and/or he was still harassing her (56.7%); (2) The provisions were too weak or punishment of violations and enforcement of the order were weak (28.9%); (3) She did not believe her partner/ex-partner would respect the order (23.3%); and (4) She perceived the court to be against her (13.3%). As Table 6 indicates, fewer rural women reported that the protective order process was good compared to the women in the urban groups, yet the majority of all groups rated the process as good.

Discussion

This is one of the first studies to examine a wide spectrum of individual, relationship, and protective order characteristics for a large and culturally diverse group of women that had obtained protective orders. Overall, women in this sample who had obtained protective orders were in their early thirties with two children, which is consistent with the findings of other studies examining demographic characteristics of women who obtain or file for protective orders against violent intimate partners (Gist et al. 2001; Linares et al. 1999). A significant proportion of women reported having less than a high school education and no GED (18–38%) and being unemployed (40–70%). Twenty to 30% reported income below the poverty threshold for a single household with two dependents (U.S. Census 2000), and being financially dependent on their DVO partner. These socioeconomic characteristics have a major impact on how women obtain resources to leave violent relationships, especially if they become the primary caretaker for their children, which is a common living arrangement after divorce (Bryson and Casper 1998). Obtaining a protective order is considered a proactive step in securing safety for partner violence victims. However, a common complaint from many court personnel, judges, and helping professionals is that women often do not pursue court action for violations of the order (Logan et al. 2005). The economic hardship of women with protective orders may help explain the lack of follow-through with the protective order process. If women do not have other help or resources to leave an abusive partner, they may need to return to their partners, which may affect their willingness or ability to follow through with enforcement of the order or may compromise their credibility with the courts (Logan et al. 2004).

The finding that women with protective orders experienced high levels of physical, psychological, and sexual violence is consistent with other literature (Carlson et al. 1999; Gondolf et al. 1994; Keilitz et al. 1997; Klein 1996; Ptacek 1999). High proportions of all three groups of women had experienced psychological and physical violence. Further, about 20% of all three groups had experienced sexual assault and more than one-third to one-half of the groups had experienced sexual insistence. While sexual assault rates were not significantly different across the three groups, rates of sexual insistence were different, with significantly fewer African American women reporting sexual insistence compared to rural women. It is not clear whether sexual insistence is a distinct part of an abuse continuum or if it should be classified as sexual assault. Clearly, the dimensions of sexual behavior within violent relationships need more research (Cole et al. 2005; Logan et al. 2007). Results from this study may suggest that a more in-depth examination of sexual behavior in violent relationships is needed that also considers cultural norms, in part, because these results potentially suggest that definitions of sexual aggression within violent relationships may be culturally bound.

Further, rural women had longer relationships with the violent partner than the other two groups and the vast majority of rural women had been or were married to that partner. In addition, a greater proportion of rural women had children in common with the DVO partner compared to the urban groups. The rural women also reported fewer economic resources and options than women in the two urban groups. Moreover, more rural women had experienced various types of abuse in the relationship with their DVO partner. These relationship and economic differences are significant in terms of the legal issues that may arise in the process of obtaining a protective order. As mentioned above, ongoing custody and legal issues can ensue from filing a protective order, which means women need adequate financial resources to successfully navigate this process.

Also, it should be noted that results of this study suggest that it is not common for custody issues to be addressed with protective orders and, even when they are awarded custody through the protective orders, women are not usually ordered child support. Leaving the issue of custody and child support unaddressed potentially exacerbates the economic and court involvement circumstances for married and cohabitating women. This is an important concern for all three groups of women; however it seems even more salient for the rural women who appear to have the fewest economic resources. Further, the combination of economic and child custody aspects of the legal process may pose a significant barrier for women in obtaining safety protections. Protective orders are an important avenue for women to pursue to protect themselves from violent partners. However, it is not clear how effective they are over the long-term especially when there are other issues to be resolved through ongoing court proceedings (e.g., divorce, custody, and child support). Moreover, the results indicate that rural communities may need a wider array of social, economic, and other services to help women in obtaining safety from violent partners.

When examining the protective order stipulations, the rural women reported that the court system granted more stipulations when compared to the urban White or urban African American women. In fact, differences in protective order stipulations for rural women were expected given differences in local court systems (Logan et al. 2005); however, differences in stipulations between the two urban groups were unanticipated. The finding that no-contact orders and stipulations for women’s partners to participate in offender counseling were less common among African American women when compared to urban White women is noteworthy. It is not clear whether the courts perceive orders for perpetrators to enter counseling as less necessary for African American women because of their shorter duration of relationships and higher rates of cohabitation or whether other unknown factors are considered. Even so, African American women did not differ significantly on key relationship variables from the urban White women. Yet rates of perpetrator-ordered counseling differed quite dramatically between the two groups. It is possible that the difference in referrals of perpetrators into counseling is due to court views of the suitability of African American men to offender treatment, but this requires further study. The differences in justice system responses regarding protective orders strongly illustrates the serious need for research to explain the factors that result in these differences.

With conflicting literature about the overall value of protective orders (Buzawa and Buzawa 1996), it is important to understand how women victims perceive the effects of orders. About a quarter of women in all three groups indicated their protective order had been violated, but these interviews were conducted relatively soon after the issuances of the order. The rate of protective order violations has been reported in the literature as ranging between 23 and 70% (Carlson et al. 1999; Harrell and Smith 1996; Keilitz et al. 1997; Tjaden and Thoennes 2000). It is interesting that there were no group differences in violation rates. Also, many women in the study agreed that the protective order process was good and that they believed the order was effective. Other studies have found that women report life improvements, feeling better about themselves, and feeling safer after obtaining a protective order (Harrell et al. 1993; Keilitz et al. 1997). While almost 70% of the women in this study rated protective orders as effective, about one-third of the sample indicated that either they did not believe the protective order was effective or they were unsure of its effectiveness. Importantly, these views of effectiveness were reported very soon after the issuance of the protective order when there would have been only limited experience with the protections. A smaller proportion of women indicated they did not feel safe or free from their partner.

Findings suggest that these women may need more support and additional resources to cope with the abuse to ensure their own and their children’s safety. Overall, however, the results of this study suggest that, even though the protective order process has problems including barriers to obtaining and enforcing terms and stipulations of the order (Logan et al. 2006), they are an important avenue of protection and potential empowerment for women experiencing partner violence. Further research should clarify the elements of protective orders that are effective over time, and should also identify the barriers or problems in obtaining, implementing, and enforcing protective orders.

This study has limitations that should be considered. The study examined women who obtained protective orders shortly after obtaining the order in only one state. A full understanding of protective orders will need to incorporate information from women who do and do not obtain orders from a variety of jurisdictions, and should examine women in different cultural contexts over time to examine trajectories of violence and enforcement of orders. Another limitation in this paper is the definition of cultural group. Although geographic area and race are often used as proxies for culture across a wide diversity of literature, race/ethnicity and geographic living area are likely only approximations of culture (Kaplan and Bennett 2003). Future research is needed to better understand how to measure and define cultural groups particularly where partner violence is concerned because culture may influence norms and behavior. However, for this paper, given the potential differences in court implementation, the regional context expanded the potential for examining how protective orders policies and procedures may differ across jurisdictions even in one state even though the jurisdictions are under the same laws. In addition, the information presented in this study was descriptive and self-reported.

Even with the limitations, this study contributes to the literature as one of the first studies to offer extensive descriptive information about relationship characteristics and perceptions of women from different cultural contexts who have obtained protective orders. Results suggest that the urban White and the urban African American women in this sample were very similar with regard to self-reported relationship and socioeconomic characteristics as well as on perceptions of protective order effectiveness and satisfaction. This finding was unanticipated given the literature suggesting that African American women may have different experiences of partner violence because of socioeconomic factors such as poverty, limited educational opportunities, unemployment, and problems with housing (Bent-Goodley 2001). Some literature also suggests that African American women may experience difficulty in accessing services due to lack of cultural competence of service providers, lack of transportation, racism, and gender role norms within African American communities (Bent-Goodley 2001; Hampton et al. 2003). This study shows that the most dramatic differences in findings were between the rural White women and the urban women (both the White and African American women) on a variety of factors including longer involvement in the violent relationship, worse economic circumstances, more victimization, more restrictive protective order stipulations, and feeling less safe when compared to women in the urban groups. Our findings introduce a new complexity to the study of partner violence among women in different cultural contexts by examining both differences in urban versus rural environments as well as by race. Future research examining partner violence and help-seeking among women from different cultural contexts (including geographic differences) may be important to facilitate services and legal protections.