Spousal Abuse in Zimbabwe: Nature and Extent across Socio-Economic Class, Gender and Religiosity

This study investigated the nature and extent of spousal abuse among the different sexes, social classes, religious and non-religious families in Zimbabwe.  Five types of spousal abuse were considered: physical, emotional, economical, sexual and psychological. The relationships between these types of spousal abuse by gender, social class and religious affiliation were tested. A total of 130 informants (mean age=33.8 years; s.d= 6.8 years) who knew families with abuse took part in this study.  Participants provided quantitative and qualitative data on the types and prevalence of spousal abuse they perceived in families they knew. The prevalence of spousal abuse did not differ by gender. There was no significant difference in the prevalence of spousal abuse between working class and middle class families. The prevalence of spousal abuse was lower among religious families. The study contradicted the view that spousal abuse was higher among the low socio-economic groups and females.

experience Domestic violence against women has been acknowledged worldwide as a violation of basic human rights, and an increasing amount of research highlights the health burdens, intergenerational effects and demographic consequences of such violence (United Nations General assembly, 1991;Heise, Ellsburg & Gottemoeller,1998;Jejeeboy, 1998).UNHCR(2003) reported that worldwide an estimated 40-70% of homicides are committed by intimate partners and that one in every three women have been beaten, coerced into sex or otherwise abused in their life.
Studies from U.S.A, Zimbabwe and Nicaraqua have shown that abused women seek medical treatment more.They have more surgeries, more physician and pharmacy visits, more hospital stays and a plethora of mental health consultations (Central Statistics Office, 2007;Osirim, 2006cited in CSOMI,2007)) thus increasing health costs.A study of abused women in Managua, Nicaragua, revealed that abused women earned 46% less than women who did not suffer abuse, even after controlling for other factors that affect earnings.Several studies have investigated correlates of violence against women (United nations, 1995;the Musasa project, 1997;Ndlovu, 1997;WHO, 2000 andCentral Statistical Office andMacro International Inc, 2007) at the exclusion of violence against men.Contrary to popular belief, men may also be abused by their wives.
The most complete information about the extent of spousal abuse in Canada comes from the 1999 General Social Survey on Victimization (GSS).This survey addressed almost 26,000 women and men in Canada about their experiences of spousal abuse in their current or previous marriages and common law partnerships.According to the GSS, women and men experience similar rates of both physical violence and emotional abuse in their relationships.However, the survey also reported that the violence experienced by women tended to be more severe -and more often repeatedthan the violence directed at men.Spousal abuse is the use of coercive power (Cvetkovich, 1984) by husband or wife on the partner.It includes physical, sexual, emotional, psychological, economic and spiritual abuse.This researcher's primary aim was to establish the prevalence or extent of spousal abuse among the different socio-economic status groups, gender, religious and non-religious groups in Zimbabwe.

Correlates of spousal abuse
Studies show that abusers are often motivated by feelings of powerlessness and insecurity (United Nations, 1995;1993;Carson, Butcher & Coleman, 1988).Jacobson and Gottman (1998) concluded that the purpose of spousal battering is to control and intimidate.It may result from unhealthy possessiveness and/or jealousy.Spousal abuse is common in cultures that undervalue or control and oppress women (Taylor & Stewart, 1991;Musasa project, 1997;CSOMI, 2007).

Cultural practices
Zimbabwe is presently experiencing major political and economic setbacks, which could lead to an increase in domestic violence.Poverty is increasing drastically, seriously destabilizing families and influencing women's situation and their capacity to demand their rights.Cultures and religions that generally accept and promote polygamy which could lead to spousal abuse, including the practice of arranged marriages could also result in spousal abuse because of lack of consent.Similarly, payment of the bride price is also used to control and oppress women in that women may feel obliged not to question the husband's actions, plans and decisions (Omiyi, 1993;Kirimire, 1993).
Violence against women is common in patriarchal societies (Ndikimbela, 2008).

The abuse cycle: Impact on quality of life
There is a cycle of violence that often begins with a pattern of verbal denigration and emotional abuse and intensifies until it manifests itself as a form of physical abuse.
Long after the black and blue bruises and broken bones from physical abuse heal, verbal abuse continues to silently erode its victim's self-worth ( Taylor & Stewart, 1991;CSOMI,2007).

Gendered violence
Women who experienced both physical and sexual violence were more likely to seek help than women who experienced only physical or sexual violence.Research suggests that physical violence in intimate relationships is often accompanied by psychological abuse, in one third to over half the cases ( Krug, Dalhberg, Mercy, Zwi & Lozano, 2002).Canada's national survey on violence against women reported that 30% of physically abused wives had to cease regular activities due to the abuse, and 50% of women could not go to work because of the harm sustained (General Social Survey on Victimization,1999).Not much work has been done on violence against men.Early studies that merely asked "have you been a victim of domestic violence" did find far lower levels of male victims; but when they asked about specific behaviors ("have you been slapped, punched,...), the numbers evened out (www.nationmaster.com,2008).The U.S.A justice department studies show that men are 32 percent less likely than women to report any form of violent victimization (www.nationmaster, 2008).Statistically, reports of women being abused are more common than that of men hence the beliefs that man are not vulnerable.However, Fiebert (n.d) from California State University has compiled findings from research relating to spousal abuse by women on men.This research examined 155 scholarly investigations: 126 empirical studies and 29 reviews and/or analyses, which demonstrated that women are as physically aggressive, or more aggressive, than men in their relationships with their spouses.Mann(n.d) reported that 59 percent of women jailed in Indiana for spousal murder claimed self-defense and that 30 percent had previously been arrested for violent crimes.Female abuse is often retaliatory in nature (CSOMI,2007;Walker, 1979) while male abuse may be a result of jealousy, a desire to maintain control or feeling of proprietary entitlement.GSS (1999) revealed that the rate of spousal homicide was much higher for Aboriginal women than men.Research by Mills(1990) and Mould (1990) showed that equal numbers of women and men commit acts of physical aggression on their partners in Britain.However, (Frude,1994;Gates, 1978) claimed that men are in most cases bigger and stronger and more familiar with the use of violence than their female counterparts and therefore likely to perpetrate spousal abuse.

The Zimbabwe situation
The inclusion of the domestic violence module in the 2007 Zimbabwe Demographic Health Survey (ZDHS) was in recognition of the presence of gender based violence.The violence has grave consequences to children, society and the country as a whole.This study has the potential to inform interventions by health personnel, policy makers and families on curbing and understanding the occurrence of spousal abuse.In Zimbabwe no study has been done to check on the effects of Gender Based Violence (GBV) in the work place.However, Zimbabwe Republic Police has reported that more than 20 women are assaulted everyday and more than 60% of murder cases that go through the courts are a result of GBV ( ZRP log books, 2008;Afro News ,n.d).ZRP also reported that 249 women were killed because of alleged infidelity in

2007.
The few studies that have been conducted in Zimbabwe suggest that women experience alarming levels of violence, primarily by their partners (Ndlovu,1997).

Social class
The ZDHS data showed that women in four quintiles across all levels of social classes similarly experienced sexual violence, with the exception of those in the highest social class quintile who experienced lower levels (CSOMI, 2007).Education and wealth did not have a great impact on ever -experience with sexual abuse but a higher wealth status was associated with a lower occurrence of spousal violence.However, a few women were found to have abused their spouses without provocation, women who are 40-49 years(4%), employed(3%), live in urban areas(4%), in the highest quintile(4%) and women who have acquired more than a secondary education(5%).Sugarman and Hotaling(1989) revealed that spousal abuse was more prevalent among people of low socio-economic status in the U.S.A. Walker (1979) and the Ministry of Health in Norway (1984), reported that most abused spouses were from middle to higher income homes where the power of wealth was in the hands of one spouse.Straus (1979) and Levinson (1989) concluded that people with a wider status difference between partners were associated with a higher frequency of spousal abuse.This may be because the one with a low status feels stuck in the relationship for they may not be capable of looking after themselves if they decide to leave.

Religious families
Glenn and Shelton (1985) cited in Brinkerhoff and White (1988) reported that people, who are religious, live in stable neighborhoods, have networks of family and friends are more likely to have stable and happy marriages.Besham, House, Nickum, and Moses (1984) revealed that religious families had lower rates of spousal abuse than non-religious families.However, Britten and Britten (1999) concluded that even highly religious people were involved in extra marital affairs and were also abusive.

Goals of the study
The study sought to establish the prevalence or extent of spousal abuse among the different socio-economic status groups, gender and religious and non-religious groups in Zimbabwe.
The questions: Are women the major recipients of spousal abuse as compared to men?Is the prevalence of spousal abuse significantly lower among people of high socioeconomic classes?Is spousal abuse in non-religious families significantly more than that in religious families?

Participants and setting
Participants were people who self-identified to know a family in abuse, enough to report on it.Purposive sampling was used for the selection of both the middle class and the working class groups.A purposive sample of 130 people participated (females=92, males=38, mean age=33.8years,standard deviation=6.8years).They were recruited from government departments and industry (low class=56, middle class=74).The sample was drawn mainly from workers in professional and management positions(66%) and clerical staff and general workers(34%).

Procedure
Data were collected from respondents during working hours.Participants consented individually to participate in the study.Participants were informed of their right to withdraw from the study at any time without penalties.

Measures Instruments and reliability
For the pilot study, 3 different questionnaires were used to establish which one among them was the most reliable.The one with the highest reliability was used for the main study.The questionnaires were designed to collect information on the extent to which men and women have experienced physical, sexual, psychological, emotional and economic abuse.The following questionnaire was used for the study: Other Reported Spousal Abuse (ORSA)-The ORSA was aimed at establishing whether the respondent knew anyone who was being abused.The questions included were as follows a) Do you know someone who has been punched?b) Is the person you know being deprived of money?and so on.The idea of using this questionnaire was that people maybe free to report about others more honestly than they would about themselves.ORSA had a cronbach alpha reliability coefficient of r=.86 (see table 1).
Cronbach alpha was used to estimate internal consistency for the different types of abuse.The cronbach alpha reliability analysis coefficients for the different types of abuse were also calculated (Table 1).For the physical and economic abuse, observed reliability coefficients fell within the moderate range and for emotional and psychological abuse scales the observed reliability coefficients fell within the high range.Given the low reliability coefficient for sexual abuse, the scale was dropped from the analysis.

Demographics
Participants also reported on the demographics of the families they were reporting on.The demographics included spouse's educational level, age, residence, marital status, occupation and sex.

Data analysis
Data was analyzed using SPSS package version 8.0.The data for the four types of abuse (physical, economical, emotional and psychological) did not meet the assumption of normality of the distribution of dependent variables scores.Log transformed variables were used for the analysis.Tests of significance were computed for types of spousal abuse by the predictor variables (Sex, Social Class and Religiosity).
One tailed test of significance were applied, placing the full 5 % confidence interval level in one end of the tail.

Correlation among the Dependent measures
Pearson product moment correlation coefficients between the different types of abuse were calculated.Table 2 below shows the correlations among the dependent variables.Psychological and economical abuse were positively and significantly correlated (p<.01).Economic and emotional abuse were also positively and significantly correlated (p<.05).This meant people who have been subjected to psychological abuse are also likely to experience emotional and economic forms of abuse.Those who have been subjected to economic abuse are more likely to experience emotional abuse.Physical abuse was not significantly correlated to any of the other three types of abuse.In other words, people who were subjected to physical abuse were not likely to experience economic, emotional and psychological forms of abuse.The correlations were of low magnitude.This suggests that the dependent variables were reasonably independent to each other to warrant separate analysis.The results also showed that there was a high correlation between emotional, psychological and economic abuse.However, people who experienced physical abuse were not likely to experience the other forms of abuse.

Effects of Sex, Social Class and Religiosity
Table 3 presents the results for ANOVA main effects for types of abuse.The results showed that perceived spousal abuse reported by others did not differ by sex or social class (p>.05).
could precipitate conflict and abuse among spouses.Thus there might be no difference in prevalence of abuse if both classes are experiencing economic conflict.

Families with or without religious affiliation
The findings of the study suggest higher levels of spousal abuse among nonreligious as compared to religious families.According to the present study, being religious does not preclude marriages from spousal abuse.The results are consistent with statistics gathered nationwide in the U.S.A that indicate that higher rates of abuse are among families without religious affiliation ( Besham et al,1984).They are also consistent with Glenn and Shelton (1985) cited in Brinkerhoff and White(1988) who reported that people who are religious, who live in stable neighborhoods, who have networks of family and friends are more likely to have stable and happy marriages.
A limitation of this study was the method of sampling used.With purposive sampling, the validity of inferences to a population cannot be ascertained (Peldhazur & Scmelkin, 1991).Future studies could use stratified random sampling since it would meet more adequately the assumptions of the parametric inferential statistical tests that were used.Spousal abuse by others may represent the actual prevalence of abuse.Future studies should contrast the findings of this study with those from using self-report data.In conclusion, religiosity is associated with a lower prevalence of spousal abuse among Zimbabwean couples.