Correlates of Violence Against Married Women in Areas of Kenya

Correlates of Violence Against Married Women in Areas of Kenya

Correlates of Violence Against Married Women in the
Maseno and Nairobi Areas of Kenya

Japheth Jaoko


            In Kenya, it is extremely difficult to obtain the prevalence rates of violence by husbands against their wives because the problem in many cases is still accepted as a cultural practice or a private affair and thus is rarely reported to authorities.  Nevertheless, reports in the popular literature and magazines indicate that the problem exists in serious magnitudes.  This exploratory study sets out to examine the prevalence of physical, sexual, and emotional violence by husbands against their wives in the Maseno and Nairobi areas of Kenya.  The Institutional Review Board at the University of Texas at Arlington approved the study as part of the author’s doctoral studies.  Of the 208 women who participated in the study, 49.5% reported a history of violence.  Prevalence rate for physical violence was 45.7%, sexual violence 12.9%, and emotional violence 39.4%.


Violence against women is a serious violation of human rights.  This worldwide phenomenon is perpetrated against women in many forms, such as intimate partner violence, sexual violence, trafficking, forced prostitution, physical and sexual violence against prostitutes, sex selective abortion, female infanticide, deliberate neglect of girls, and rape in war.1  Other forms of violence against women include inhumane cultural and religious practices, such as female genital mutilation, dowry killing in India, and Middle Eastern honor murders.2

            Violence against women is a serious global health problem associated with a range of health concerns such as physical injuries, emotional and psychological complications, sexually transmitted diseases, and pregnancy complications.3  In addition, a World Bank report pointed out that violence on women accounts for as much health problems and death in women as cancer, and surpasses both road accidents and malaria combined as a cause for health problems for women.4  This phenomenon transcends national boundaries, geographic locations, socioeconomic status, culture, religion, and ethnicity.5

            Although violence against women has received increased scientific, public, and policy attention over the last  two  decades, there is still a paucity of scientific research on the problem in Kenya.  A review of the literature on domestic violence reveals that to date, no national or community-based scientific study has been carried out to examine the prevalence of physical, sexual, and emotional violence against married women in Kenya.  Although no national prevalence statistics for Kenya are available, the Kenyan Chapter of the International Federation of Women Lawyers (FIDA) estimates a prevalence rate of about 50%.6

            In an effort to fill a crucial knowledge gap, this exploratory study sets out to examine the prevalence of physical, sexual, and emotional violence committed against women within a marriage relationship in the Maseno and Nairobi areas of Kenya.  For the purposes of this study, violence is defined as any act perpetuated by the husband that is harmful to the wife.  This includes physical attacks, threatened physical attacks, psychological or emotional hostility, sexual assaults or threatened sexual assaults, and neglectful conducts.

Literature Review

            Over the last two decades, the evidence of the scope and significance of violence against women has increased  throughout most of the world.  In the United States, estimates of the prevalence of violence against women suggests that nine hundred thousand to three million women suffer abuse at the hands of a partner each year.7

            In Europe, numerous national studies on the prevalence of violence have been conducted with varying results.  The prevalence estimates range from 6% in Switzerland 8 to 26.2% in the Netherlands.9

            In Africa, Asia, and the Middle East, it is very difficult to get scientific information on the prevalence of violence against women, as it appears to be extremely under-reported.  Specifically, in sub-Saharan Africa, empirical evidence is limited to a small number of studies.  The prevalence estimates for South Africa is 46%.10  In Tanzania, the reported rates are 45.7% for verbal abuse, 37.6% for physical abuse, and 16.3% for sexual abuse.11  Due to under- reporting, common in many survey studies, it would be safe to assume that the true prevalence estimates in most countries are probably higher than what is reported above.

Theories of Domestic Violence

            Numerous causal theories have been developed to explain violence against women.  These theories can be broadly placed into six categories as follow: 1) biological  2) psychological, 3) systems theory, 4) sociological, 5) social structural, and 6) nested ecological model.

            Biological  models  focus upon  three  general  approaches  to understanding the etiology of violence against women: organic, genetics, and neurochemical.12  For example, head injuries that result in brain infections or other changes in the structure of the brain have been reported to put a person at an elevated risk for becoming violent.13

            Psychological theories, like the biological models, focus on individual factors.  However, unlike biological models, psychological theories do not focus on organic causes.  Rather, the focus is mainly on individual personality traits or on childhood and other past experiences that are assumed to shape individuals to become violent in adult relationships.  Factors such as stress, hopelessness, self- esteem, childhood abuse, and childhood exposure to violence are often analyzed under this perspective.14

            Systems theory locates the causes of violence against women within the nuclear family unit.  Within this theory, wife abuse is viewed as a problem within the family unit dynamics rather than just as a problem with only the abusive husband.  Violence is viewed to take place within unique relational contexts in which all family members interact to maintain the family system.  Wife abuse is thus regarded as a family dysfunction, with both the husband and wife contributing to the problem.15

            Sociological theories of domestic violence shift the focus from micro-level to macro-level analyses.  Instead of looking at individual factors in isolation, the focus is widened to include factors in the families of origin, societies, and cultures.  A key approach under the sociological theories is the social learning theory.16

            When applied to the sphere of domestic violence, social learning theory postulates that observing and experiencing violence in the family of origin and/or within a broader society is a contributing risk factor to being violent.17  This perspective is commonly referred to as the intergenerational transmission of violence when applied to family violence.

            Similar to sociological  theories, social structural theories of domestic violence analyze the etiology of wife abuse at a macro-level.  This perspective postulates that wife abuse is rooted in the structure of societies rather than the organic pathologies or personality traits of individual men.18  A key approach that fits under the category of social structural theory is feminist theory.  The feminist approach attempts to explain domestic violence by focusing on gender analysis of power.19  Wife battering, in this case, is seen as a method through which men seek to control women.

            The nested ecological model considers a wide variety of factors that interact at different levels within a broader social environment to cause family violence.20  It attempts to analyze the etiology of violence against women at four different levels: ontogenetic, microsystem, exosystem, and macrosystem.  The ontogenetic level focuses on risk factors that increase the likelihood of domestic violence at the individual level.  These factors may include biological or organic factors, individual personality traits, or individual developmental experiences that are assumed to shape one to become violent.  Within the microsystem level, the focus is on relationship patterns within a family, family stressors, and structural factors of the family.  The exosystem level focuses on factors at the community level.  For example, unavailability of jobs or a steady income 21 and high levels of community crime have been found to be risk factors for domestic violence.22  The macrosystem level focuses on factors at the societal level such as societal cultural beliefs and values in relation to domestic violence.  For example, wife abuse in South Africa may be viewed as a reflection of the country’s widespread culture of violence that affirms violence as a conflict resolution method.23


            This study employed a non-experimental, ex-post facto design to measure the prevalence of physical, sexual, and emotional violence against married women in the Maseno and Nairobi areas of Kenya.  Data for the study was collected via a quantitative survey conducted using a non-probability, purposive sampling procedure.  A total of 208 married women participated in the survey.

            The incidence and prevalence of physical, sexual, and emotional violence were measured using the Partner Abuse Scale: Physical (PASPH) and the Partner Abuse Scale: Non-Physical (PASNP).  Both the PASPH and the PASNP were developed by Walter Hudson.24

            The starting point for recruiting participants consisted of visiting local women groups at local universities, churches, market places, and neighborhoods throughout Maseno and Nairobi areas of Kenya.  Four trained female research assistants contacted potential participants on behalf of this author and those who agreed to participate were asked to complete the survey.  Those who agreed to participate in the study helped to recruit additional participants from among their acquaintances.  The survey was presented as an anonymous and voluntary study, with no names or identifying information requested.

Data Analysis

            Three levels of statistical analysis were conducted on the data from this study.  First, an initial univariate analysis of the data was performed, with descriptive statistics conducted to summarize the data in a manageable form.  Second, bivariate analysis was conducted on each predictor variable and criterion variable to determine the individual effects of the predictor variables on the criterion variable.  Lastly, multivariate analysis was conducted using logistic regression in order to detect the presence or absence of the criterion variable based on a set of predictor variables.

            With a dichotomous criterion variable, chi square tests were performed in cases where the predictor variables are categorical variables and t tests were performed in cases where the predictor variables are continuous.  All predictor variables that were found to be significant were included in the multivariate analysis.

            Multivariate analysis was conducted using logistic regression in order to detect the presence or absence of physical abuse based on the various predictor variables.  Logistic regression was deemed most appropriate due to the use of a dichotomous criterion variable as well as the use of both continuous and categorical predictor variables.


            Participants in the study ranged in age from 19 to 64, with a mean age of 33.48 years.  Participants’ husbands were slightly older, ranging in age from 21 to 70, with a mean age of 36.75 years. Participants had been married an average of 8.75 years, with a range of 1 to 41 years of marriage.  The median number of children per family was 2, with a range of 0 to 8 children per family.

            Participants’ level of education ranged from 4.8% having no formal education to 7.2% having a graduate level education.  Almost a third, 28.8%, of participants reported that they had a secondary education. Concerning participants’ husbands, 2.4%  had no formal education with 9.1% having a graduate level education.  Over  a fourth, 27.8%, of participants’ husbands had a secondary education.  Concerning employment status, 64.9% of participants were employed, compared with 87.5% of their husbands.

            Of the 208 women interviewed, 49.5% (n=103) reported having suffered some form of violence from their husbands.  Prevalence rate for physical violence was 45.7% (n=95), sexual violence 12.9% (n=27), and emotional violence 39.4% (n=82).  All the participants (100%, n=27) who suffered sexual violence also suffered emotional and physical violence.  Additionally, almost all of the participants who suffered emotional violence (90%, n=74) also suffered physical violence.  Only 3.8% (n=8) of the women suffered emotional violence without accompanying physical or sexual violence.

            Participants who had experienced any form of violence had significantly less education (M = 3.46) than did participants who had not experienced violence (M = 4.42).  Similarly, participants who had experienced violence had husbands who had significantly less education (M = 3.73) than did participants’ husbands who had not experienced violence (M = 4.83).

            Additionally, almost six times as many participants, 61.1%, who had experienced violence were not employed.  This figure compares to 10.6% of participants who had not experienced violence who were not employed.  Furthermore, almost seven times as many, 23.2 %, of participants who had experienced violence reported that their husbands were not employed.  This figure compares to 3.5% of participants who had not experienced violence who reported that their husbands were not employed.  No significant associations were found between wife abuse and the number of years of marriage and the number of children in a household.

Summary and Conclusion

            In this study, almost half of all the women participants reported having suffered some form of violence from their husbands.  However, because of the private and sensitive nature of domestic violence in many areas of Kenya, it would be safe to assume that the true prevalence estimates are probably higher than what is reported in this study.

            The findings of this study supports those of previous studies that reported that women with low educational attainments are at a higher risk for domestic violence,25 and that males with lower levels of education are more likely to be abusive to their female partners.26  However, other studies have reported that there is no significant association between risk of violence and either a woman’s educational attainment or a male partner’s educational attainment.27

            This study also found that unemployed women were at a greater risk for domestic violence than those employed.  Other studies have reported that a woman’s access to independent income increased her risk of experiencing domestic violence,28 while yet others have reported that the effect of a woman’s employment status  is insignificant to her risk of violence.29  Further cross-national studies are needed to examine the factors behind the seemingly contradictory findings.

            Consistent with the findings of this study, unemployment of a male partner has been reported elsewhere to be associated with increased risk of violence on women.30  Previous studies have also shown that employment in low-status jobs increases the odds of male to female violence.31

            Although the findings of this study cannot be generalized to all of Kenya, the findings do suggest that a large number of women in Maseno and Nairobi areas are suffering violence in the hands of their husbands.  This phenomenon seems to be pervasive even though the Kenya Government is a signatory to the United Nations’ Convention on the Elimination of All Forms of Discrimination Against Women (CEDAW).32  As a signatory to that convention, the Kenyan government is obligated, among other things, to eliminate violence against women in all its forms.

            To make any progress in reducing  the prevalence of violence against women in Kenya, significant changes in governmental and societal attitudes towards domestic violence must be undertaken.  Strategies to address the needs of abused women in the short term and to combat the root causes in the long term must be instituted.  Public educational awareness programs must be developed to sensitize the citizens to the  fact that violence on women is dehumanizing, illegal, and has serious health consequences.  Additionally, provision of shelters and support services for abused women must be made available.

            Although more research is certainly warranted to study violence against Kenyan women beyond Maseno and Nairobi areas, perhaps the findings of this study could play a modest role towards the efforts to liberate abused Kenyan women from the bondage of violence they suffer silently in the hands of their male partners.  All Kenyan women deserve to feel safe in their homes.


            1 C. Watts and C. Zimmerman, “Violence Against Women: Global Scope and Magnitude,” Lancet 359:1232-1238, n9313, 2002.

            2 S. K. Araji, “Family Violence Including Crimes of Honor in Jordan,” Violence Against Women 7:876-900, n5, 2001.

            3 R. Jewkes, “Violence Against Women: An Emerging Health Problem,” International Clinical Psychopharmacology 15:37-45, n3, 2000.

            4 World Bank, World Development Report: Investing in Health (New York: Oxford University Press, 1993).

            5 M. A. Ellsberg, R. Peha, A. Herrera, J. Liljestrand, and A. Winkvist, “Wife Abuse Among Women of Childbearing Age in Nicaragua,” Journal of Public Health 89:241-245, n2, 1999.

            6 FIDA (Kenya), “Report of a Baseline Survey Among Women in Nairobi,” March 2002.

            7 United States Department of Justice (1998). Violence Against Intimates: “Analysis of Data on Crimes by Current or Former Spouses, Boyfriends, and Girlfriends,” March 1998.

            8 L. Gillioz, “Domination and Violence Against Women Within the Couple,” European Strategies to Combat Violence Against Women (Copenhagen, Denmark: WHO Regional Office for Europe, 1997).

            9 R. Romkens, “Prevalence of Wife Abuse in the Netherlands,” Journal of Interpersonal Violence 12:119-125, n1, 1997.

            10 Z. Dangor, L. A. Hoff, and R. Scot, “Woman Abuse in South Africa: An Exploratory,” Violence Against Women 4:125-152, n2, 1998.

            11 S. Maman, M. D. Sweat, J. K. Mbwambo, N. M. Hogan, G. P. Kilonzo, J. C. Campbell, and E. Weiss, “HIV-Positive Women Report More Lifetime Partner Violence: Findings from a Voluntary Counseling and Testing Clinic in Dar-es-Salaam, Tanzania,” American Journal of Public Health 92:1331-1338, n8, 2002.

            12 J. Alper, “Biological Influences on Criminal Behavior: How Good is the Evidence?,” British Medical Journal 310, 372, 1995.

            13 H. Johnson, “Violence and Biology: A Review of the Literature,” Families in Society 77:3-18, n1, 1996.

            14 M. Straus, R. Gelles, and S. Steinmetz, Behind Closed Doors: Violence in the American Family (New York: Doubleday, 1980).

            15 K. Healey, C. Smith, and C. O’Sullivan, Batterer Intervention: Program Approaches and Criminal Justice Strategies. Issues and Practice in Criminal Justice (Washington: National Institute of Justice, 1998).

            16 A. Bandura, Social Learning Theory (Englewood Cliffs, NJ: Prentice- Hall, 1977).

            17 S. Mihalic and D. Elliott, “A Social Learning Theory Model of Marital Violence,” Journal of Family Violence 12:21-47, n1, 1997.

            18 K. Yllo, “Through a Feminist Lens: Gender, Power, and Violence,” R. J. Gelles and D. R. Loseke, eds., Current Controversies on Family Violence (Newbury Park, CA: Sage Publications, 1993), pp. 47-62.

            19 E. Pence and M. Paymar, Education Groups for Men Who Batter: The Duluth Model (New York: Springer, 1993).

            20 D. DePanfilis, “Intervening with Families When Children are Neglected,” H. Dubowitz, ed., Neglected Children: Research, Practice, and Policy (Thousand Oaks, CA: Sage, 1998), pp. 211-236.

            21 G. Kaufman-Kantor and M. Straus, Report on the USAF Family Needs Screener (Durham, NH: Family Research Laboratory, University of New Hampshire, 1999).

            22 A. Andrew, “Developing Community Systems for the Primary Prevention of Family Violence,” Family and Community Health 16:1-9, n4, 1996.

            23 L. Vogelman and G. Eagle, “Overcoming Endemic Violence Against Women in South Africa,” Social Justice 18:209-229, 1991.

            24 W. Hudson, WALMYR Assessment Scales and Scoring Manual (Tallahassee, FL: WALMYR Publishing Co., 1997).

            25 V. Rao, “Wife Beating in Rural South India: A Qualitative and Econometric Analysis,” Social Science and Medicine 44:1169-1180, 1997

            26 S. Martin, A. Tsui, K. Maitra, and R. Marinshaw, “Domestic Violence in Northern India,” American Journal of Epidemiology 150:417-426, 1999.

            27 S. Arscott-Mills, “Intimate Partner Violence in Jamaica: A Descriptive Study of Women who Access the Services of the Women’s Crisis Center in Kingston,” Violence Against Women 7:1284-1302, n11, 2001.

            28 S. R. Schuler, M. Syed, and S. H. Badal, “Men’s Violence Against Women in Rural Bangladesh: Undermined or exacerbated by microcredit programmes?,” Development in Practice 8:148-158, n2, 1998.

            29 T. Angless, M. Maconaschie, and M. V. Zyl, “Battered Women Seeking Solutions: A South African Study,” Violence Against Women 4:637- 658, n2, 1998.

            30 R. Castro, C. Peek-Asa, and A. Ruiz, “Violence Against Women in Mexico: A Study of Abuse Before and During Pregnancy,” American Journal of Public Health 93:1110-1117, n7, 2003.

            31 G. L. Fox, M. L. Benson, A. A. DeMaris, and J. V. Wyk, “Economic Distress and Intimate Violence: Testing Family Stress and Resources Theories,” Journal of Marriage and Family 64:793-808, n3, 2002.

            32 United Nations, Convention on the Elimination of All Forms of Discrimination Against Women (New York: United Nations General Assembly, 1979).