Need for a gender-sensitive human security framework: results of a quantitative study of human security and sexual violence in Djohong District, Cameroon
© Parmar et al.; licensee BioMed Central Ltd. 2014
Received: 24 October 2013
Accepted: 23 April 2014
Published: 7 May 2014
Human security shifts traditional concepts of security from interstate conflict and the absence of war to the security of the individual. Broad definitions of human security include livelihoods and food security, health, psychosocial well-being, enjoyment of civil and political rights and freedom from oppression, and personal safety, in addition to absence of conflict.
In March 2010, we undertook a population-based health and livelihood study of female refugees from conflict-affected Central African Republic living in Djohong District, Cameroon and their female counterparts within the Cameroonian host community. Embedded within the survey instrument were indicators of human security derived from the Leaning-Arie model that defined three domains of psychosocial stability suggesting individuals and communities are most stable when their core attachments to home, community and the future are intact.
While the female refugee human security outcomes describe a population successfully assimilated and thriving in their new environments based on these three domains, the ability of human security indicators to predict the presence or absence of lifetime and six-month sexual violence was inadequate. Using receiver operating characteristic (ROC) analysis, the study demonstrates that common human security indicators do not uncover either lifetime or recent prevalence of sexual violence.
These data suggest that current gender-blind approaches of describing human security are missing serious threats to the safety of one half of the population and that efforts to develop robust human security indicators should include those that specifically measure violence against women.
Human security: background
Human security shifts the traditional concept of security from interstate conflict and the absence of war to the security of the individual . In doing so, it recognizes that ‘the basic rights of people, not merely the absence of military conflict between states, are fundamental to world stability ’. Despite that clarity, concepts of human security vary greatly. Some authors advocate for a broad definition, including threats to livelihoods and food security, health, psychosocial well-being, enjoyment of civil and political rights and freedom from oppression, and personal safety, in addition to absence of conflict. While inclusive, this broad definition can be unwieldy and difficult to operationalize [3, 4]. King and Murray emphasize freedom from violence, advocating for a narrower definition that includes those factors which are ‘important enough for human beings to fight over or to put their lives or property at great risk’, and identify five key indicators: poverty, health, education, political freedom, and democracy . However, as pointed out by Paris in 2001, narrower definitions risk excluding important threats to personal security that do not relate to these five indicators . Hastings developed a Human Security Index that attempts to move beyond United Nations Development Programme’s (UNDP) Human Development Index to include various national measures of stability and inequity in society (including gender, environment, health, a ‘social fabric index’, and measures of corruption). His approach, however, does not include a measure of violence against women, a measure he suggests is a ‘women’s health issue’ as opposed to a human security issue . Authors have posited that a ‘gender-blind’ approach to human security will similarly miss important threats to individual and community security [7–9].
Multiple authors have explored the role of inequality and oppression in inter- and intrastate violence. In general, studies have found that greater social disparities between segments of society are associated with greater levels of conflict [8, 10, 11]. The role of gender inequity in conflict has also been previously explored. Melander found that states with higher rates of gender inequity, as measured by female participation in parliament and female to male higher education attainment ratio, have higher rates of intrastate conflict . Similarly, Caprioli found that states with higher levels of gender inequity had higher rates in intrastate violence between 1960 and 2001 .
Assessing levels of human security remains challenging. Given the links described between human security indicators and a greater propensity to intra- and interstate conflict, measurement of human security could be used both to herald potential deterioration within a conflict and to monitor progress of post-conflict state conditions. Leaning and Arie in 2001  developed a model designed for this purpose. Based on experiential knowledge and extensive qualitative review of human security as understood in African populations that have been insecure, the Leaning-Arie model argues that a population requires a ‘core bundle of basic resources: material psychological and social’ to ensure a minimum level of survival. ‘If minimum material inputs can be guaranteed and if efforts can be made to shore up basic social coping capacities, societies will be more stable and less prone to fragmentation, violence and atrocity’ . This model moves beyond measurement of basic material supports (food, water, shelter, income) to incorporate critical psychosocial components associated with individual security: namely, ‘a sense of home and safety, constructive family and social supports, and an acceptance of the past and positive grasp of the future ’. These three domains of home, community, and positive sense of future reflect the ways humans relate to themselves, each other, and the world: ‘[humans] are anchored and seek safety in a sense of home; they find meaning, identity, and functional support in their relationships with family and community; and they build their lives through time, depending upon their sense of trust in the future and their sense of freedom from the past ’. The main strength of this model is its ability to detect threats to human security at a more granular level that may be used to predict community level violence—a key difference from other approaches that focus on aggregate national indicators. This model, however, does not explicitly incorporate gender sensitive measures.
The following study explores human security and gender using quantitative measures in Djohong District, Cameroon, a rural region to where members of the Mbororo tribe from the Central African Republic (CAR) have fled in response to a decade of targeted killings, kidnappings, forced taxation and extortion, beatings, and the burning and looting of their homes and villages [13, 14]. These human rights violations have occurred at the hands of government troops, anti-government rebel groups, and opportunistic gangs of bandits known locally as coupeurs de route. At the time of this study, an estimated 80,000 had fled across CAR’s western border into Cameroon , 21,000 of whom, as of 2010, lived as refugees in the sparsely populated and mainly agricultural Djohong. During their flight, the coupeurs de route targeted the Mbororo, locally perceived as wealthy pastoralists, stealing their livestock or kidnapping their family members and forcing them to pay large ransoms. Arriving in Djohong with little or no assets, the Mbororo refugees had become considerably dependent on the international aid community and Cameroonian host population. In early 2007, the United Nations Office for the Coordination of Humanitarian Affairs (UNOCHA) cited a high burden of sexual violence in the internally displaced and refugee population, but detailed statistics were lacking .
We examined human security indicators based on the Leaning-Arie model and performed a population-based study for measures of sexual and gender based violence. From this, we used the model to determine whether a human security framework that does not specifically incorporate measures related to gender would capture insecurity faced by women affected by displacement and recent conflict, and to assess whether or not gender-specific measures are needed when measuring human security.
In March 2010, we undertook a population-based study conducted in Djohong District, Cameroon to better ascertain women’s health needs within the refugee and Cameroonian host communities . Embedded within the survey instrument were indicators of human security derived from the Leaning-Arie model, assessing three domains of psychosocial stability that suggest individuals and communities are most stable when their core attachments to home, community and the future are intact. The Leaning-Arie model was selected as it was designed to be used for community-level, ongoing measurement of human security that could be used to assess a community’s tendency toward instability and conflict over time . The domain of sense of attachment to home includes a ‘sustainable sense of home and safety providing identity, recognition, and freedom from fear’ measured as 1) number of years in current location; 2) proportion of the population who own land; and 3) self-reported sense of safety. Sense of attachment to community, the second domain in the model, elaborates the individual’s relationship to their community as ‘a network of constructive social or familial support, providing identity, recognition, participation and autonomy’ , such that sudden changes in economic or political circumstances, or perceived inequalities relative to others that appear not to be improving can potentially precipitate violent behavior. Indicators used to measure attachment to community include 1) measures of dynamic inequality between members of different groups, including income and assets; and 2) subjective sense of ties to and inclusion in one’s community or village. The third domain is a positive grasp of the future which includes ‘the confidence that one’s own capacities, and the external social and political structures that confer meaning and stability, will persist for some indefinite period into the future’ . This can be measured by an individual’s plans to remain in the current village, raise children in the current village, and plans to plant for future harvests. The text below outlines these three domains and provides examples of how each was specifically studied.
Human security framework: Leaning Arie model
Attachment to Home:
Number of years living in this village
Does any extended family live in this village?
Do you own land?
Do you feel safe in this village?
Attachment to Community:
Is your home made from mud brick or grass/reeds/sticks/tarp?
How much land do you own? How much livestock do you own?
Do you feel attached to the community in this village?
If you needed medical care, would someone take you to the hospital?
Positive outlook for the Future:
Do you want to grow old and die in this village?
Will your children be working in this village in 10 years?
Will you be using your land in 2 years?
Do you plan to stay in this village?
Female refugee and host population participants were selected using a two-stage household cluster design. The first stage included a random selection of 40 clusters (villages) weighted by refugee and host population estimates from the UN High Commissioner for Refugees (UNHCR) registration lists of January 2010 and from Djohong District government officials, respectively. The latter figures were 2010 estimates of the Cameroonian population using 2006 census figures adjusted for annual population growth. Since the human security dataset came from a study designed to also characterize gender-based violence (GBV) in the region, the estimated prevalence of GBV was used to determine sample size. Based on an estimated prevalence of 20-30% from previous GBV studies in Africa , a desired precision of 0.05, and a design effect of 2.0, we estimated a sample size between 500 and 650, assuming a non-response rate of 5%. Since observations within a cluster may be more alike than observations across clusters, particularly with shared perceptions of human security, we took this intra-cluster correlation into account in the sample size calculation. This design effect, defined as the ratio of the variance taking into account the cluster sample design and variance of a simple random sample design with the same number of observations, was conservatively estimated at 2.0 based both on previous sexual violence studies [19, 20]. To minimize the possibility of increased intra-cluster correlation and homogeneity on the outcome of sexual violence prevalence and to limit design effect, we sampled more clusters in the first stage and fewer households in the second stage ; the second stage included 15 randomly selected women per cluster for a final sample size of 600 respondents.
In stage one, clusters were defined as villages. To account for the modest variation in village populations across the district, probability proportional to size (PPS) sampling was used to randomly select clusters. As the refugee population does not live in segregated areas but is interspersed among the host Cameroonian villages , we reasoned that random sampling in each village would result in a sample demographic similar to the population demographic. All villages in Djohong District were considered for possible selection. The World Health Organization’s (WHO) Extended Program on Immunization (EPI) method  was used in the second stage to randomly select the 15 respondent households in each village. A household was defined as a group of individuals living under the same roof and eating meals from the same pot. Village chiefs assisted the research team to identify the geographic center of each village selected. From that point, the team randomly selected a direction by spinning a pen on a flat surface and then randomly selected a number of houses to pass to reach the first sampled household. Each subsequent household whose door was nearest to the door of the previous household was surveyed until all 15 surveys within the cluster were completed. To minimize non-response, a pre-visit announcement was sent to each village cluster to request the presence of all adult women in the village for the day of sampling. Three attempts were made to contact selected households where respondents were initially unavailable. When, as a result of PPS sampling, larger villages contained more than one cluster, clusters were geographically distributed according to the location of population centers.
At each village cluster, the study was announced as a women’s health study for purposes of safety and confidentiality. The team queried the adult female (≥18 years of age) head of household on household demographics, human security, household economy and assets, level of education, food security, water, fuel, shelter, access to health care, self-reported mental health, self-reported reproductive health and sexual violence experienced by the respondent only. In polygamous households, the senior wife was interviewed, a decision based on cultural custom.
For the purposes of this study, sexual violence was defined as any physically or verbally forced sexual act, including molestation, forced undressing, forced or unwelcome touching of a sexual nature, forced intercourse, forced insertion of an object into any body cavity or any other non-consensual sexual act, whether completed or not.
Perpetrators were defined as the person or persons responsible for the forced act. Refugees were defined according to the UNHCR definition: ‘owing to a well-founded fear of being persecuted for reasons of race, religion, nationality, membership of a particular social group or political opinion, is outside the country of his nationality, and is unable to, or owing to such fear, is unwilling to avail himself of the protection of that country ’. Given considerations of confidentiality, no identification or proof was used to verify refugee status.
The survey instrument, much of which had been field tested the previous year for a baseline prevalence study on sexual violence, was translated and back-translated by bilingual French-English speakers in Cameroon, then tested and colloquially adjusted by data collectors who were local health professionals (with the Cameroonian Ministry of Health at Djohong Hospital) as well as local community health workers on staff with an international health non-governmental organization (NGO). Training included detailed explorations of each question with fine-tuning of terms and translation in French and in the local language, Fulfulde. Data collectors were then trained on the sampling methodology, including the EPI method, using simulated models of villages with varying configurations. Following training on the survey instrument, interviews were conducted in Fulfulde in a setting that ensured privacy and confidentiality.
Human subjects protection
This study was conducted in accordance with the WHO ethical and safety recommendations for researching, documenting and monitoring sexual violence in emergencies . Due to high rates of illiteracy, the team obtained oral consent prior to administration of the survey and specifically informed the respondents that the survey would ask about sexual violence as a women’s health and livelihood concern. Prior to survey questions on sexual violence, a second verbal consent was obtained to allow the participant an opportunity to refuse to answer questions in this section. Respondents were assured that their names would not be recorded, that there would be no penalties or benefits for refusing or agreeing to participate, and were offered access to counseling and medical services through the international NGO and local providers as needed. The study was reviewed and approved by the Office of Human Research Administration at the Harvard School of Public Health. All male data collectors were required to have previous experience in the care of female survivors of sexual violence, either as clinicians or as sexual violence counselors. While WHO guidelines recommend female surveyors and translators whenever possible, the lack of female staff and the availability of experienced sexual violence male counselors in the region necessitated their inclusion as surveyors. The survey instrument did not identify respondents.
The lead field investigators checked the data for errors, then coded and entered it daily into a password protected Excel spreadsheet. Non-identifiable hard copies were stored in a locked facility on the NGO compound in Djohong District. At the end of the survey, the lead investigators securely transported hard copies to a locked storage facility at the Harvard Humanitarian Initiative where they remained accessible only to the lead investigators. STATA 11 (StataCorp, College Station, TX) was used for the analysis of the imported spreadsheet. Cluster sampling design was accounted for in the analysis; a generalized Hansen-Hurwitz estimator for a two-stage cluster design was used to estimate the means and percentages and confidence intervals were constructed by calculating the standard error of the generalized Hansen-Hurwitz estimator. Receiver operating curve (ROC) analysis was used to evaluate the predictive value of human security indicators for both lifetime and six-month sexual violence. Three predictive models were analyzed for their ability to predict sexual violence within the refugee and host populations: 1) age and ethnicity; 2) age, ethnicity, time in village, and all human security indicators; and 3) age, ethnicity, time in village and community human security indicators only.
Refugees n = 191
Cameroonian n = 397
Proportion of respondents*
32.5 (23.3 – 41.7)
67.5 (58.3 – 76.7)
35.1 (32.8 – 37.4)
34.5 (32.7 – 36.2)
Mean household size
6.6 (6.0 – 7.2)
5.9 (5.6 – 6.2)
Sense of attachment to home
Human security outcomes: sense of attachment to home
Refugee (n = 190)
Refugee % (95% CI)
Cameroonian (n = 396)
Cameroonian % (95% CI)
Years in village
3.9 (3.2 – 4.5)
17.9 (15.7 – 20.1)
Does your extended family live in this village?
44.7 (36.8 – 52.7)
59.8 (52.8 – 66.9)
Do you own land?
68.3 (56.4 – 80.1)
87.6 (81.8 – 93.3)
I feel safe in this village
98.4 (96.8 – 100)
95.7 (93.4 – 98.0)
Sense of attachment to community
Human security outcomes: sense of attachment to community, ethnicity
Refugee (n = 191)
Refugee % (95% CI)
Cameroonian (n = 397)
Cameroonian % (95% CI)
55.5 (43.8 – 66.6)
9.8 (6.3 – 15.1)
14.1 (9.3 – 21.0)
53.9 (40.2 – 67.1)
10.5 (5.5 – 19.0)
20.9 (13.3 – 31.4)
11.0 (5.5 – 20.9)
5.5 (3.2 – 9.5)
3.1 (1.0 – 9.5)
2.5 (0.9 – 6.4)
5.8 (2.8 – 11.6)
7.3 (3.6 – 14.2)
83.3 (74.2 – 89.6)
47.9 (36.2 – 59.7)
16.8 (10.4 – 25.8)
52.1 (40.3 – 63.8)
Human security outcomes: sense of attachment to community, home
Refugee (n = 191)
Refugee % (95% CI)
Cameroonian (n = 396)
Cameroonian % (95% CI)
Water comes from a protected well
56.0 (43.2 – 68.9)
34.6 (22.9 – 46.3)
Home is a made from mud brick
95.3 (90.6 – 100)
98.0 (96.5 – 99.5)
Home is made of grass, reeds, sticks or tarp
4.7 (0 – 9.4)
2.0 (0 – 3.5)
Human security outcomes: sense of attachment to community, assets
Refugee mean (95% CI)
Cameroonian mean (95% CI)
Average daily income, CFA
842.38 (593.58 – 1091.18)
1172.53 (1001.66 – 1343.39)
Average size of land owned, hectares
1.54 (0.38 – 2.69)
1.26 (0.68 – 1.83)
Cows owned by household
6.6 (0.2 – 12.9)
6.2 (3.9 – 8.5)
Chickens owned by household
2.3 (1.5 - 3.2)
2.3 (1.6 – 3.0)
Day without food in the past month
1.7 (1.6 – 1.8)
1.7 (1.7 – 1.8)
Though confidence intervals overlap, Cameroonian households had a higher average daily income (refugee 842.38 CFA, 95% CI 593.58 – 1091.18 vs Cameroonian 1172.53 CFA, 95% CI 1001.66 – 1343.39). Refugees had access to protected wells in greater proportions (56.0% 95% CI 43.2 – 68.9) than the host population (34.6% 95% CI 22.9 – 46.3) (Table 4).
Human security outcomes: sense of attachment to the community, perceptions in the affirmative
Refugee % (95% CI)**
Cameroonian % (95% CI)**
If I was alone in this village someone would help
39.6 (29.4 – 49.7)
56.2 (49.5 – 62.9)
If I needed money, someone in this village other than family would help
30.2 (20.5 – 39.8)
47.0 (39.6 – 54.3)
If I needed urgent medical care, non-family villagers would take me to the hospital
40.4 (29.5 – 51.4)
59.7 (52.3 – 67.1)
I feel attached to the community in this village
90.6 (86.1 – 95.1)
90.9 (86.2 – 95.6)
The members of my community form strong relationships and rely on each other for support
70.5 (61.3 – 79.8)
75.5 (68.0 – 83.0)
Positive grasp of the future
Human security outcomes: sense of future
Refugee % (95% CI)**
Cameroonian % (95% CI)**
I want to grow old and die in this village
83.2 (77.7 – 88.8)
81.5 (76.6 – 86.5)
My children will live in this village in 10 years
83.2 (77.3 – 89.0)
76.3 (70.8 – 81.9)
My children will work in this village in 10 years
74.2 (66.3 – 82.1)
63.5 (55.5 – 71.5 )
I will use my land in 2 years
98.1 (97.2 – 99.8)
98.5 (96.0 – 100)
I plan to stay here
91.5 (86.8 – 96.2)
91.0 (87.8 – 94.2)
Female heads of household were asked whether or not they had experienced sexual violence during their lifetimes or during the past six months. A complete report on the prevalence of sexual violence in this population has been previously published ; a selection of these data are presented here to highlight the context of human security measures.
Sexual violence, total population
# with characteristic
Weighted % (95% CI)*
35.2 (28.7 – 42.2)
64.0 (54.3 – 72.5)
Friend or Member of the Community
20.0 (14.2 – 27.5)
Soldier/rebel or coupeurs de route***
17.1 (10.7 – 26.1)
2.9 (1.2 – 7.2)
Estimated population-wide sexual violence by partner/husband
22.3 (17.1 – 28.5)
Sexual violence, broken down by Refugee vs Cameroonian
Refugee n = 77
Refugee % (95% CI)*
Cameroonian n = 128
Cameroonian % (95% CI)*
40.8 (31.5 – 50.8)
32.4 (25.2 – 40.5)
Sexual violence in the past 6 months
39.0 (21.7 – 59.6)
38.3 (28.4 – 49.2)
52.0 (35.4 – 68.1)
71.1 (61.0 – 79.5)
Member of my family
Friend or Member of the Community
15.6 (8.3 – 27.2)
21.2 (13.9 – 32.8)
Soldier/rebel or coupeurs de route***
39.0 (25.6 – 54.2)
3.9 (1.4 – 10.5)
3.9 (1.3 – 11.5)
2.3 (0.7 – 7.3)
37.7 (22.9 – 52.5)
6.3 (1.6 – 10.9)
Among the 206 surveyed female heads of household who reported an incident of sexual violence during their lifetime, 64% (95% CI 54.3 - 72.5) reported that the perpetrator was their husband or partner. For refugee female heads of households, 52.0% (95% CI 35.4 - 68.1) of sexual violence was perpetrated by a husband or partner compared to 71.1% (95% CI 61.0 - 79.5) for Cameroonian female heads of households.
The second most commonly identified perpetrators were soldiers/rebels or coupeurs de route. Of the 206 respondents who reported at least one incident of sexual violence during their lifetimes, 17.1% (95% CI 10.7 - 26.1) reported this type of perpetrator. Furthermore, this perpetrator was more commonly reported by refugee women, 39.0% (95% CI 25.6 - 54.2) compared to the host population 3.9% (95% CI 1.4 - 10.5). A total of eight female refugee heads of household reported multiple perpetrator sexual violence by soldiers/rebels and coupeurs de route.
Of the 206 female heads of household who reported sexual violence, 41 (20.0%, 95% CI 14.2 - 27.5) experienced sexual violence perpetrated by a friend or member of their community. Refugee survivors of sexual violence described the perpetrator as a friend or member of their community 15.6% (95% CI 8.3 - 27.2) of the time, while the affected female host population reported the perpetrator as friend or member of their community 21.2% (95% CI 13.9 - 32.8) of the time. Among refugees who reported sexual violence, 37.7% (95% CI 22.9 - 52.5) reported multiple perpetrators, while 6.3% (95% CI 1.6 - 10.9) of the host population who survived sexual violence reported multiple perpetrators.
A majority of both refugee and host population sexual violence survivors had not reported these events to the authorities. Only 24.7% (95% CI 13.8 - 40.1 of refugee women and 6.3% (95% CI 3.1 - 12.4) of host population sexual violence survivors had reported this sexual violence to any authority. Fourteen women surveyed, including eight refugees and six Cameroonians, reported ‘punishment’ for their perpetrators. These ‘punishments’ included being forced to marry the victim, payment, and in only three cases incarceration .
Predictive value of human security indicators for gender-based violence
Refugees and host population report similar levels of human security in terms of attachment to home, community, and a positive grasp of the future. Though some human security indicators differ between these populations, overall they reflect a relatively stable population at the time of this study, where refugees not only feel safe but also have access to land, livelihoods, clean water, shelter, and wish to remain in their newly adopted villages for many years to come. Refugees appeared to be able to attain stability and security in Eastern Cameroon—in fact, though the proportion of refugees that owned land was lower than the proportion of host population that owned land, refugees and Cameroonians who owned land owned in similar amounts, suggesting equal opportunity to grow livelihoods between the two groups. Similarly, refugees had equal levels of assets as many host population households. Refugees and Cameroonians were both similarly attached to their villages, desiring to stay in the same village well into the future.
However, despite the unusual and encouraging stability and egalitarian quality of these human security findings, these human security indicators missed an epidemic of sexual violence that endangered the lives and health of refugee and host population women. ROC analysis shows that human security indicators measured in this study did not uncover either lifetime or six-month sexual violence. These data suggest that current, gender-blind means of describing human security are missing serious threats to the safety of one half of the population. If human security is ‘concerned with the protection of people from critical and life-threatening dangers’ , this presents a major gap.
Why is it crucial that measures of human security be sensitive to gender? A robust body of evidence has explicitly linked measures of gender inequity and violence against women, including intimate partner and domestic violence, to higher levels of conflict [8, 12, 26–29]. Much of this work has been done by Caprioli, whose analyses have shown that increased domestic gender equity has a pacifying effect on state behavior at an international level , and that states with higher levels of gender equality are less likely to resort to violence first in resolving international disputes . Caprioli went on in 2005 to link lower levels of gender equity to higher rates of internal conflict , and more violent conflict . Melander reproduced these findings, demonstrating that societies with higher female representation in parliament and higher female to male education attainment ratios have lower levels of internal conflict . There are several theories as to why this may be true. Some suggest that a tendency toward peaceful behavior is associated with a tolerance of the rights of others , while others cite that states with a tendency toward inequity and oppression are inherently likely to be more violent both internally and internationally [8, 11]. Violence against women not only has implications for the human security of half of the world’s population, it has consequences for the stability of the state.
Human security indicators represent current feelings of respondents, and thus are not likely to be sensitive to distant events. Lifetime sexual violence could have occurred any time over the respondent’s lifetime, and in either CAR or Cameroon, making it difficult to interpret the meaning of high human security indicators in the context of a high prevalence of lifetime sexual violence. However, human security indicators were also not associated with any meaningful association with sexual violence during the past six months, suggesting an insensitivity of this model to a major threat to the security and health of women and girls. Further, lifetime and six month prevalence for sexual violence reported by both women in the CAR and Cameroon were similar, and both proportions were substantial. Sexual violence represents one extreme manifestation of gender inequity—this level of violence is likely to be associated with other manifestations, including limited access to education and access to justice. Low levels of literacy, low rates of completion of primary and secondary school education, poor access to justice for survivors of sexual violence, and high rates of early marriage were outlined in a previous publication focused on sexual violence findings of this study . Thus, sexual violence against women represents a proxy indicator—it suggests that gender inequity has real human security consequences in this population, consequences that are not captured by this quantitative model, and consequences that extend beyond sexual violence itself.
These data also confirm what was described in the report of the Human Security Research Group published in 2012 . This report outlined data suggesting that rape as a weapon of war, while important in many contexts, does not constitute the majority of rape faced by women in conflict. The findings of this report support this assertion [Tables 8 and 9], as the majority of women who experienced sexual violence reported that the perpetrator was their intimate partner (64.0%, 95% CI 54.3-72.5) and the second most common perpetrator was a friend or member of the community (20.0%, 95% CI 14.2-27.5). The group identified as ‘combatants’ - soldiers, rebels, and coupers de route— who, in most conflict contexts, would be the focus of sexual violence related insecurity—constituted a comparatively less commonly reported perpetrator (17.1%, 95% CI 10.7-26.1). When disaggregated data is examined [Tables 8 and 9], data show that while refugees report a higher proportion of sexual violence at the hands of combatants (39%, 95% CI 25.6-54.2) than Cameroonian women, refugees still report that a majority of the perpetrators of sexual violence are known to them (intimate partners: 52%, 95% CI 35.4-68.1: friends/community members: 15.6%, 95% CI 8.3-27.2). Based on these findings, we estimate 22.3% (95% CI 17.1-28.5) of all women in Djohong District, Cameroon have endured sexual violence by their intimate partner during their lifetimes. Several authors have reported high rates of sexual violence in times of conflict [31–34], and as suggested by the Human Security Research Group, much of that occurs at the hands of intimate partners and community members. While a focus on sexual violence perpetrated by armed forces remains important, the much more common epidemic of violence perpetrated by intimate partners and community members against women must also be addressed, particularly given that this violence presents risks not only for the affected women but for the development and maintenance of their communities and the world as a whole.
Several measures have been developed to understand gender equity at the country level, including UN Women’s Indicators and Statistics Database, GenderStats, and the World Economic Forum’s Gender Gap Project, and many studies analyze other measures including female to male education attainment ratio, female representation in parliament, life expectancy, literacy, and participation in democracy. However as Hudson points out, these do not include measures of physical violence against women . These are also aggregated sources at the country level, and not likely to aid in a nuanced understanding of community level human security, as is the goal of the Leaning and Arie model.
Based on the findings of this study, we suggest that, at a minimum, measures of human security should include prevalence of sexual violence at the hands of all types of perpetrators, including both combatants and non-combatants. Additionally, measures of gender inequality that have been previously linked to conflict in other research (female participation in parliament and female to male higher education attainment ratio) should be included in human security frameworks.
All interviews were conducted privately in their homes in the respondent’s preferred language. Nuanced understanding of questions at the household level could be distorted in the course of translation and risked being asked differently in Fulfilde across data collectors. Though all efforts were made to ensure the respondents privacy at the time of the interview, it is possible that women felt reluctant to report sexual violence given strong social stigma associated with sexual violence. Whenever possible, respondents were interviewed by women, however, several members of the surveying team were male, due to limitations in availability of experienced multilingual staff. Although the few male data collectors had prior experience in medical and psychological intake of sexual violence survivors, it is possible that women felt less comfortable reporting their experiences with sexual violence to them regardless. We proceeded with male data collectors based on the acceptance in this community of male sexual violence counselors in a pre-existing international NGO program on sexual violence. Although female data collectors are recommended according to WHO guidelines , several studies have successfully engaged male data collectors in similar research [19, 33]. As all data reported here refer to the respondent’s last episode of sexual violence, it is likely that these statistics under-represent the true burden of sexual violence in this community. Results are specific to refugee and host population female heads of household over the age of 18 only, thus results cannot be generalized to all women in the region and are likely biased towards the experiences of older women. Given these female heads of household tend to be first wives in polygamous households, this study represents the experiences of this group of women primarily. Arguably any form of sexual violence could affect human security and data was not collected on men or children.
Several of the data collectors were NGO sexual violence program staff and thus known to be affiliated with these programs. Although interviewers were carefully trained to emphasize that no aid or compensation would be given for participation in the survey, it is possible respondents may have altered their responses. Respondents may either have reported themselves to be refugees in the interests of obtaining services, or Cameroonians in order to avoid stigma. Given the anonymous nature of this survey and initial disclosure regarding aid and compensation, we do not feel it is likely that this issue significantly impacted responses.
Human security indicators were developed by the research team based on the Leaning-Arie model, used during a 2009 iteration of a population based survey in the region and refined based on the results. Outcomes are self-reported, thus assets, income, and land holdings were not verified objectively.
This study suggests that in order to move toward an effective, accurate, and predictive quantitative measure of human security, measures of sexual and gender based violence must be included. A gender-inclusive model would not only ensure accurate measurement of what remains one of the main threats to women’s security, namely violence against women and girls both as a result of war and as a result of rape and physical violence in their homes and communities, but it would ensure a more sensitive means of understanding a community’s and state’s propensity toward violence. This will ultimately not only move us toward increasing the safety of women and girls, it could ultimately lead to a more secure and peaceful world.
The research team wishes to acknowledge the substantive input of Dr. Jennifer Leaning on developing the indicators that define the human security model.
- UNDP: Human Development Report. Book Human Development Report. 1994, Oxford University PressGoogle Scholar
- Axworthy L: A new scientific field and policy lens. Secur Dialog. 2004, 35: 348-349. 10.1177/096701060403500308.View ArticleGoogle Scholar
- Paris R: Human security: paradigm shift or hot air?. Int Secur. 2001, 26: 87-102. 10.1162/016228801753191141.View ArticleGoogle Scholar
- McDonald M: Human security and the construction of society. Global Soc. 2010, 16: 277-295.View ArticleGoogle Scholar
- King G, Murray C: Rethinking human security. Polit Sci Q. 2000, 116: 585-610.View ArticleGoogle Scholar
- Hastings DA: From Human Development to Human Security: A Prototype Human Security Index. Book From Human Development to Human Security: A Prototype Human Security Index. 2009, United Nations Economic and Social Commission for Asia and the PacificGoogle Scholar
- Blanchard EM: Gender, international relations, and the development of feminist security theory. Signs. 2003, 28: 1289-1312. 10.1086/368328.View ArticleGoogle Scholar
- Caprioli M: Primed for violence: the role of gender inequality in predicting internal conflict. Int Stud Q. 2005, 49: 161-178. 10.1111/j.0020-8833.2005.00340.x.View ArticleGoogle Scholar
- Harcourt W: Literature Review on Gender and Fragility. Book Literature Review on Gender and Fragility. 2009, European Report on DevelopmentGoogle Scholar
- Leaning J, Arie S: Human security: a framework for assessment in conflict and transition. HCPDS Working Paper Series. 2001, 11 (8): 1-73.Google Scholar
- Fearon JD: Governance and Civil War Onset. Book Governance and Civil War Onset. 2010, Stanford UniversityGoogle Scholar
- Melander E: Gender equality and intrastate armed conflict. Int Stud Q. 2005, 49: 695-714. 10.1111/j.1468-2478.2005.00384.x.View ArticleGoogle Scholar
- UNHCR chief puts spotlight on Central African refugees in Cameroon.http://www.unhcr.org/4b9108f26.html.
- CENTRAL AFRICAN REPUBLIC-CAMEROON: CAR refugees in Cameroon fear returning home.http://www.irinnews.org/report/75567/central-african-republic-cameroon-car-refugees-in-cameroon-fear-returning-home.
- Security Council Report, 2010. Central African Republic historical chronology.http://www.securitycouncilreport.org/chronology/central-african-republic.php?page=all&print=true.
- CAR: Struggling to Undo the Damage of Sexual Violence.http://www.irinnews.org/printreport.aspx?reportid=77552.
- Parmar P, Agrawal P, Greenough PG, Goyal R, Kayden S: Sexual violence among host refugee population in Djohong District, Cameroon. Global Publ Health. 2012, 7 (9): 974-994. 10.1080/17441692.2012.688061.View ArticleGoogle Scholar
- Moreno C, Jansen HAFM, Ellsberg M, Heise L, Watts C: WHO Multi-country Study on Women's Health and Domestic Violence against Women. Book WHO Multi-country Study on Women's Health and Domestic Violence against Women. 2005, Geneva SwitzerlandGoogle Scholar
- Johnson K, Asher J, Rosborough S, Raja A, Punjabi R, Beadling C, Lawry L: Association of combatant status and sexual violence with health and mental health outcomes in postconflict Liberia. JAMA. 2008, 300: 676-690. 10.1001/jama.300.6.676.View ArticlePubMedGoogle Scholar
- Fanslow J, Robinson E: Physical injuries resulting from intimate partner violence and disclosure to health providers: results from a New Zealand population-based study. Inj Prev. 2011, 17: 37-42. 10.1136/ip.2009.026104.View ArticlePubMedGoogle Scholar
- Bilukha O: Old and new cluster designs in emergency field surveys: in search of a one-fits-all solution. Emerg Themes Epidemiol. 2008, 5: 1-7. 10.1186/1742-7622-5-1.View ArticleGoogle Scholar
- Bennett S, Woods T, Liyanage WM, Smith DL: A simplified general method for cluster-sample surveys of health in developing countries. World Health Stat Q. 1991, 44: 98-106.PubMedGoogle Scholar
- Convention and Protocol Relating to the Status of Refugees.http://www.unhcr.org/3b66c2aa10.html.
- Organization WH: WHO ethical and safety recommendations for researching, documenting and monitoring sexual violence in emergencies. Book WHO ethical and safety recommendations for researching, documenting and monitoring sexual violence in emergencies. 2007, WHOGoogle Scholar
- Thakur R: A political worldview. Secur Dialog. 2004, 35: 347-348. 10.1177/096701060403500307.View ArticleGoogle Scholar
- Caprioli M: Gendered conflict. J Peace Res. 2000, 37: 53-68.View ArticleGoogle Scholar
- Caprioli M: Gender equality and state aggression: the impact of domestic gender equality on state first use of force. Int Interact: Empir Theor Res Int Relat. 2003, 29: 195-214.View ArticleGoogle Scholar
- Caprioli M, Boyer MA: Gender, violence, and international crisis. J Conflict Resolut. 2001, 45: 503-518. 10.1177/0022002701045004005.View ArticleGoogle Scholar
- Hudson V, Caprioli M, Ballif-Spanvill B, McDermott R, Emmett CF: The heart of the matter: the security of women and the security of states. Int Secur. 2009, 33: 7-45. 10.1162/isec.2009.33.3.7.View ArticleGoogle Scholar
- Group HSR: Human Security Report 2012, Sexual Violence, Education, and War: Beyond the Mainstream Narrative. Book Human Security Report 2012, Sexual Violence, Education, and War: Beyond the Mainstream Narrative. 2012, Simon Fraser University; Human Security Research GroupGoogle Scholar
- Frasier PY, Belton L, Hooten E, Campbell MK, DeVellis B, Benedict S, Carillo C, Gonzales P, Kelsey K, Meier A: Disaster down east: using participatory action research to explore intimate partner violence in Eastern North Carolina. Health Educ Behav. 2004, 31: 69S-84S. 10.1177/1090198104266035.View ArticlePubMedGoogle Scholar
- Anastario M, Sheheb N, Lawry L: Increased GBV among women internally displaced in Mississippi two years post hurricane Katrina. Disaster Med Public Health Prep. 3: 18-26.Google Scholar
- Johnson K, Scott J, Rughita B, Kisielweski M, Asher J, Ong R, Lawry L: Association of sexual violence and human rights violations with physical and mental health in territories of Eastern democratic Republic of the Congo. JAMA. 2010, 304: 553-562. 10.1001/jama.2010.1086.View ArticlePubMedGoogle Scholar
- Amowitz L, Reis C, Lyons KH, Vann B, Akinsulure-Smith AM, Taylor L, Iacopino V: Prevalence of war-related sexual violence and other human rights abuses among internally displaced persons in Sierra Leone. JAMA. 2002, 287: 513-521. 10.1001/jama.287.4.513.View ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.