Women’s perceptions of effects of war on intimate partner violence and gender roles in two post-conflict West African Countries: consequences and unexpected opportunities

The influence of cultural and social norms on IPV

Certain cultural and social norms play a particularly important role in creating conditions that support or facilitate IPV[5]. These include the belief that a man has the right to assert power over a woman and is socially superior; that a woman’s freedom should be restricted; that physical violence is an acceptable way to resolve conflict within a relationship; and that divorce brings shame to the family[5].

Many of these norms are common in parts of sub-Saharan Africa, resulting in IPV being widely justified by both men and women as a normal part of an intimate relationship[6, 7]. In many African settings, the majority of both men and women have been found to agree that a husband has the right to use violence in response to women’s transgression of traditional gender roles, such as if she is disobedient, fails to perform household and child care duties, or is unfaithful[8–10]. Indeed, a study in rural Kenya found that while women did seek, and sometimes receive, support for IPV from informal community resources, the prevailing opinion was that this violence within partnerships was a generally accepted aspect of local culture[11]. A study analysing the Ugandan DHS data found that the majority of women (73%) and men (57%) believed that wife beating is acceptable in at least one of the following scenarios: if the wife goes out without informing her husband; if she neglects the children; if she argues with the husband; if she refuses to have sex with the husband; or if she burns the food[12]. In one specific community in Sierra Leone, the belief that a woman’s prayer goes to God through a man creates a situation in which every woman must be affiliated to a man if she is to find acceptance in the community[13]. While this particular belief is not common across all communities, it is illustrative of the complex interactions between gender roles, spirituality, and social norms that can affect women's well-being.

The transgression of gender norms and failure to fulfil cultural expectations of good womanhood and successful manhood are among the most important triggers for IPV[2]. For example, Mann & Takyi[14] write that frequently-cited reasons for marital violence relate to women behaving in a way unexpected of their gender, such as neglecting household chores, disobeying husbands or elders, and withholding sex from partners. Many have suggested that violence occurs when men are unable to live up to the socially presented ideas of what it means to be a ‘successful’ man[15, 16]. Olayanju et al.[6] note that certain aspects of life in sub-Saharan Africa make IPV more likely, including widespread poverty that makes it difficult for men to achieve what is expected of them socially.

In addition, there is some evidence that the process of cultural change, especially shifts in gender roles and expectations, is related to increased IPV. The WHO[17] cited the ‘sanctions and sanctuary’ framework, which suggests that IPV may be highest in societies where the status of women is in a state of transition: where women have a very low status, violence is not needed to enforce male authority, and where women have a high status, they will probably have achieved sufficient power collectively to change traditional gender roles. It is the middle ground where violence can spike. Studies of the short-term effects of micro-credit programmes aimed at empowering women may provide helpful examples of these transition periods. They have found mixed results in terms of the impact on IPV[18], with some finding that economic empowerment protects women from IPV whilst others report that women’s adoption of non-traditional roles creates tension in the home and precipitates IPV.

The influence of conflict on IPV

Whilst the relationship between IPV and conflict remains under-studied, there is evidence that IPV has increased during periods of conflict in Afghanistan, Lebanon, Palestinian territories, Sri Lanka, and Côte d’Ivoire[19–22]. In Côte d’Ivoire, Hossain and colleagues documented that IPV was the most frequently reported form of violence following the period of armed conflict (20.9% among women; 9.9% among men)[22]. Vinck and Pham[23] detail the evidence in support of a hypothesis that there is a relationship between conflict and IPV, including studies that have found higher rates of IPV amongst war veterans compared to civilian couples, and they trace the relationship between war experiences, subsequent mental health problems and increased likelihood of either perpetrating or experiencing IPV.

There are a number of pathways through which war experiences could potentially contribute to IPV. Annan & Brier[24] note the role played by structural factors that permit and sustain IPV in war-affected areas. In conflict settings, kinship support systems may become fractured as people are displaced; judicial systems and institutionalised conflict-resolution mechanisms no longer function effectively, and a culture of impunity may develop[25]. Although these factors are not unique to such settings, they are undoubtedly exacerbated by prolonged conflict.

The experience of conflict affects almost every aspect of socio-economic life. Poverty is amplified, and there are often limited opportunities to generate income; these stresses have the potential to contribute to IPV[26]. In a study of refugees from nine different countries, poverty was cited repeatedly as a cause of violence[27]. A man’s inability to fulfil the traditional ‘provider’ role was said to contribute to feelings of anger and shame, which manifest themselves in violence, especially when his wife or children ask for things he is unable to provide.

A loss of social role and status is a common consequence of war and displacement, as people lose their jobs, access to land for farming, and their roles within their communities[28–30]. This is particularly true for men. The nature of gendered division of labour in most societies is that men’s roles and masculinity are tied to participation within the public realm, while women’s roles and femininity are linked to the personal or household sphere. Women are often able to maintain their traditional roles and responsibilities, and so maintain their gender identity, whilst men are more likely to lose their central roles, especially that of protector and provider, which makes it difficult to maintain their sense of themselves as men in the community and family[28, 31]. Some have found connections between these losses and a tendency towards marital conflict, especially where women take on the role of breadwinner[25–27, 32, 33].

A society that lives with conflict and war may be more likely to develop social norms in which violence is seen as a normal way to resolve problems, and, therefore, in which IPV is more likely. Uthman et al.[7], in their study of 17 sub-Saharan African countries, found that women were more likely to experience IPV in societies where the use of violence is a socially accepted norm. The WHO[17] states that the normative use of violence to resolve conflicts is one of two specific risk factors that have been found to be strongly associated with IPV (the second is the unequal position of women in a particular relationship and in society).

One of the few studies of the relationship between experience of war and IPV (focusing only on physical violence) was conducted by Vinck & Pham[23] in Liberia. Their national multi-stage cluster survey of 4501 adults (2196 of whom were female) found exposure to war-related violence to be associated with increased likelihood of intimate-partner physical violence. Men who had direct experience of war-related potentially traumatic events, or who had witnessed or taken part in the conflict, were more likely to perpetrate IPV. Vinck and Pham suggest that this may reflect the impact that violence has on individuals’ socio-economic characteristics and on their belief and value systems. The qualitative study reported in this paper builds on the work of Vinck and Pham, amongst others, by exploring in some depth the relationship between conflict and IPV from the perspective of women who experienced the wars in Liberia and Sierra Leone and were living in post-conflict communities.

Conflict in Sierra Leone and Liberia

Sierra Leone is located on the west coast of Africa, bound on the north by Guinea, the south-east by Liberia and the south-west by the Atlantic Ocean. The population in 2012 was estimated to be close to 5.5 million[34].

Sierra Leone has experienced political instability since the late 1960s, but protracted armed conflict began in 1991 when the Revolutionary United Front tried to topple the constitutionally elected government by launching a guerrilla war from neighbouring Liberia. During the conflict, tens of thousands died and more than 2 million people were displaced, either to another part of Sierra Leone, into a neighbouring country or further afield. The war was characterised by mutilations and forced recruitment of children, as well as killings, sexual assault, looting and destruction of property. The Lome Peace Accord was signed in July 1999, the first dispatch of United Nations soldiers arrived soon after, and disarmament of rebel forces began the following year. In January 2002, President Kabbah declared that the civil war was officially over.

There has been no recent study of the prevalence of IPV in Sierra Leone, but in 1998 66.7% of 144 women surveyed reported having been beaten by an intimate partner[35]. Sierra Leone passed a domestic violence act in 2007, establishing basic rights for women in the home and entitlements for survivors such as free medical care, but implementation of the act has been limited[36].

Liberia is on the coast of West Africa, bordering Guinea, Ivory Coast and Sierra Leone. It has a population of approximately 4 million (estimated in 2009)[37]. The country experienced a long and divisive civil war, beginning in 1989 and ending with the signing of a peace treaty in 2003. It is estimated that more than 250,000 people were killed in fighting or massacres and one million people (one third of the population) were displaced[1]. Towns and cities were devastated, and most of the country’s infrastructure was destroyed.

Studies among the Liberian population have found high levels of IPV[38]. A survey of 4526 mothers found an incidence rate of 40% for emotional and physical IPV combined[39], and prevalence rates of 31.5-50.1% were found for physical IPV amongst sub-groups of Liberian women[40]. A national multi-stage cluster survey found that 37.7% of women reported lifetime exposure to intimate partner physical violence and 24.4% reported incidence of intimate-partner physical violence over a one-year recall period[23].

In this paper we report on women’s perceptions of the causes of IPV in West Africa and the ways in which they understand these causes to interact with the experiences of war.

Participants

IRC staff in each location recruited participants from the women who were receiving support services from the IRC. The criteria used to select participants for focus group discussions were that participants should be women over 18 and living in the target community, and that there should be diversity in terms of age and socio-economic status. They were not selected according to whether or not they had experienced IPV. Participants in Freetown and Monrovia lived in an urban setting, and those from Voinjama and Kailahun were from a rural setting.

Focus group discussions

The purpose of the focus group discussions was to explore the nature of IPV in that community, beliefs about causes and effects, options available to women experiencing IPV, and women’s help-seeking behaviour. Women were asked to share their observations of the situation in their community, rather than their own personal experiences, though in many cases women chose to share their own experiences of violence.

The term ‘domestic violence’ was commonly used in both Sierra Leone and Liberia, so was used during interviews and focus group discussions. However, we specified that we were asking about violence from an intimate partner (husband or boyfriend), rather than from other family members.

The facilitator followed a semi-structured focus group guide, with a small number of questions used to structure the discussion. These questions were not asked directly, but were explored through a broader discussion of relevant issues, including through scenarios and examples, often presented by the participants themselves.

Fourteen focus group discussions were conducted: five in Freetown, four in Kailahun, three in Monrovia and two in Voinjama. All groups except two consisted of 8 women; the total number of participants was 110. Some participants were members of IRC-facilitated Women’s Action Groups. We interviewed two groups of professional women (one in Freetown and one in Kailahun) and one group of women working in the local market (in Kailahun); the others were general members of the community who were not grouped according to any specific criteria. Most discussions took place in a Women’s Centre, with some being conducted in the IRC office and one in a school classroom. The length of the interviews ranged from 66 to 88 minutes (mean = 78 minutes).

Individual interviews

Women survivors of IPV were interviewed to explore individual experiences of IPV and their opinions on the causes of specific perpetrators’ behaviour. Interviews began with a broad question which enabled women to recount their experiences in their own way (e.g. ‘tell me the story of the violence you experienced and the impact it has had on your life’). Depending on the response to this opening question, the interviewer probed to explore the issues of interest, which included the woman’s perception of the triggers for her partner’s violence. The interview was semi-structured, with the interviewer exploring certain themes in a fluid and dynamic way and following up on any other issues of interest raised by the interviewee.

Twenty individual interviews were conducted: five in Freetown; five in Kailahun; four in Voinjama; and six in Monrovia. Women living in both urban and rural settings were interviewed, and their ages ranged from 18 to 49 (mean age = 30 years). Some were married and had experienced violence over many years; some were single and had been beaten by their boyfriends regularly; and three were single and had experienced a single very severe assault at the hands of their boyfriends.

The interviews were conducted in a location chosen by the woman. Sometimes this was the IRC office; in other cases it was a friend’s house, the IRC vehicle, a Women’s Centre, or the woman’s workplace. The length of the interview ranged between 17 and 68 minutes (mean = 38 minutes).

Translation and transcription

Our procedure for translation followed the example of Pavlish[42] in which the lead researcher asked questions in English and a translator repeated the question in the local language. The participant responded in their local language, and the translator immediately translated that response into English. The speech of the translator was recorded, and the lead researcher took her own notes in English.

The recordings were transcribed after the interview by the lead researcher and then deleted from the voice recorder and the computer on which they were played. All participants were asked whether they were comfortable with the discussion or interview being recorded, and none expressed any concern.

Analysis

Analysis was conducted by the lead researcher in two stages. Whilst in the field, a rapid analysis was conducted, in which the researcher identified key themes in the data based on the research questions. She then reviewed the data under each of these themes and identified the key issues within each theme. The issues were summarised to present during debrief meetings which took place in each location. IRC staff, research participants and other community members took part in the debrief meetings, during which the initial findings were discussed and explored. This process enabled the lead researcher to check her initial analysis, learn from the discussion of the initial findings, and ensure that it reflected participants’ understanding of the key issues.

Once the data collection stage was over, analysis was conducted using complete data from all locations, both the fully transcribed interview data and feedback obtained during the debrief meetings. The thematic analysis undertaken for the initial analysis was used to define a set of conceptual codes that was then systematically applied to all text using Nvivo9. New codes were added as they emerged, and the initial list of codes was revised until it accurately reflected the data set. The data within each code were then comprehensively reviewed, and patterns between codes explored.

Ethical issues

There were significant ethical considerations involved, and the safety and wellbeing of women was prioritised. The research was planned and conducted in accordance with the WHO’s guidelines[43] on conducting research on domestic violence. A set of safety guidelines were developed, which consisted of the following elements:

The secondary analysis of these data and publication of these findings was approved by the Institutional Review Board at Queen Margaret University, UK.

Cultural models and social norms

Whilst our findings suggest that many men and women in Sierra Leone do not adhere strictly to these traditional beliefs, the respondents said that when men did hold these beliefs strongly, IPV was more likely.

Household economic relations

Direct effects of war on IPV

Some participants in group discussions also felt that the experience of war had not affected men’s use of violence towards women because they had seen the same behaviours even before the war. However, the majority in the group discussions reported that the war did influence IPV in several ways. Women described direct effects of war on IPV, as well as how social changes related to the war affected IPV.

War-related transitions in women’s social roles

In addition, women said that after the war, many national and international NGOs started working on issues of women’s rights and empowerment and established services for women affected by IPV. A great deal of sensitisation around these issues was conducted in communities, both for men and for women.

There was less agreement about men’s responses to this change and some diversity in the experiences reported. Some women said that these changes in women’s roles and behaviours had led to an increase in IPV, whilst others said that it had led to a decrease.

Reflections on the research process

Some challenges were experienced during the data collection process, particularly regarding translation. The translators were all female IRC employees working with the Gender-Based Violence programme in the country and location in which we were conducting the research. This had advantages because they were familiar with the issues raised by the participants and the terminology used. However, they were not trained in translation, and this resulted in a tendency to summarise and use their own words rather than the participants’ words.

We must also acknowledge that all participants had some connection to IRC. Those who participated in individual interviews were receiving support from IRC. Whilst we recognise that this is a specific sub-group of women who experience IPV, for ethical reasons we choose to interview only survivors who already had access to support services. Those who participated in group discussions were not necessarily receiving direct services but were known personally or professionally by IRC staff. It may be that these women had participated in IRC awareness activities, so had more knowledge of IPV issues than other women in their communities. However, their opinions on the ways in which the conflict in their country may have affected IPV are not likely to be influenced by their connection with IRC since IRC awareness activities did not address this issue specifically.

As with all qualitative research, we cannot claim that the women who participated in this research are representative of the populations of interest, though they do reflect the diversity of those populations in terms of age, socio-economic status and ethnic group. It is worth noting, however, that since we spoke to relatively small numbers of women in each of the four locations, it is difficult to draw conclusions about whether differences between their perspectives were due to differences in the urban and rural contexts, differences between the Sierra Leone and Liberian contexts, or individual differences between our participants in the various locations. There were some indications of differences between the settings in terms of urban versus rural and country context, and future work with larger samples may help to draw these out.

An additional note on analysis is that in some studies, a second coder is used to check the reliability of the identified themes. However, in this case a single analyst coded the data, and reliability was checked by sharing the initial analysis with stakeholders in the field (members of the IRC team involved in the data collection; representatives of the participants; and other interested parties).

Implications for practice

The transitions taking place in gender relations in post-conflict Sierra Leone and Liberia have been facilitated in part by the work of national and international organisations that have been active in the country since the end of the war (or earlier) and have been building awareness around women’s rights. The rate of change varies within communities and within families, and when women’s expectations change at a different rate to their partners’, this can contribute to tension and conflict in the home. It is particularly important at this time that women have services and support available. It is also crucial that organisations understand the dynamics of this tension and conflict in the particular context and base programmes on this understanding. To accelerate the rate of change in a community before this understanding is developed, or before services and support are in place, could be very damaging.

Our findings highlight both the potential negative and positive influence that war can have on IPV and women’s rights, particularly as expressed through decision-making and economic empowerment. Given this variation, programming should focus on both preventing and remediating the negative consequences and on promoting the positive social shifts; focusing solely on the negative may lead to a missed opportunity to increase the positive social change that is already occurring in some cases in the context of conflict.

Our results suggest that one main threat to women’s safety in terms of IPV is women’s financial dependence on their partners. When we asked participants what they thought would most help women who were experiencing IPV, or would help to prevent IPV, almost all said that activities to enhance the financial independence of women would have the most impact. As we saw earlier, however, enhancing women’s independence can be threatening to men and can increase IPV. It is essential, therefore, that economic empowerment and livelihoods programming, which seek to address women’s low status and perceived inferiority, take place alongside interventions to encourage men to recognise the value of women and the benefits of women’s participation in the economy and the home. The involvement of men in activities designed to prevent gender-based violence is important, and has been recognised by IRC and other organisations (see, for example, the work of Sonke Gender Justice[48]).

Economic interventions are also important for preventing IPV among young girls from poor families who are especially vulnerable to abuse because of their total dependence on the men they marry at a young age. Enabling girls in this situation to provide for themselves (or their families to provide for them) would be a good investment of resources.

It is also necessary for support services for women who experience IPV to be reliable and accessible, both so women have somewhere to go for help and to deter men who might previously have believed they could assault their partners and ‘no-one would ask any questions’.