Gender mainstreaming in Ethiopia: translation of policy into practice and implications on the ground

 

Julie Newton[1], Wellbeing in Developing Countries ESRC Research Group[2]

 

Abstract

 

In spite of the growing recognition that ‘gender’ matters amongst development practitioners and institutions which have translated into efforts to ‘mainstream gender’, there has been an overall persistence and sometimes aggravation of gendered inequalities. This paper explores these contradictions in the context of Ethiopia. It is widely acknowledged that women in Ethiopia are disproportionately disadvantaged on a number of grounds. Drawing on data from four rural sites[3] from the Wellbeing and Developing Countries ESRC Research group, this paper makes the case for a more complex analysis of gender inequalities. It begins with an investigation of the policies and interventions in place to address gender inequalities at the national level.  It then explores how patterns of power at the community, household and individual level are inherently gendered in ways that have particularly negative effects on women.  It also examines how gender inequalities are being contested at different levels. The paper then concludes with a discussion of the effectiveness of the policies and structures in place at the national level to address gendered inequalities.

 

 

 

1. Introduction

 

The Beijing Platform of Action (1995) was hailed a significant victory for feminists globally, signalling universal recognition of the importance of bringing gender centre stage within the development agenda. The term ‘gender mainstreaming’ was subsequently introduced as the key strategy for achieving gender equality resulting in a proliferation of governments, national and international bilateral organisations and NGOs championing it as an essential development objective. Ten years on, a recent review of progress on the commitments made in the Beijing Declaration and Platform of Action presents an ambivalent record of progress and a general atmosphere of discontent and disillusionment (Molyneux & Razavi, 2005). In spite of progress for women globally on a number of development indicators, there has been an overall persistence and sometimes aggravation of gendered inequalities (Molyneux & Razavi, 2005). These are particularly pronounced within Africa in countries such as Ethiopia. This has put the success of gender mainstreaming under question causing many to argue that gender mainstreaming is in crisis. This paper explores these contradictions in the context of Ethiopia, drawing on findings from the WeD research group.

 

The paper begins by drawing on a variety of secondary sources to provide a brief summary of how women in Ethiopia are disproportionately disadvantaged on a number of grounds. It then provides an account of how gender has been mainstreamed in Ethiopia and how gender has been incorporated within key Ethiopian policy documents, particularly in relation to addressing gendered inequalities. The paper then turns to the empirical findings of the Wellbeing and Developing Countries (WeD) ESRC Research group to make the case for a more complex analysis of gender inequalities. It explores how women’s inequality is reinforced through four key dimensions: gendered division of labour, pregnancy infancy cycle, marriage and violence against women’s bodies.  It briefly examines how gender inequalities are being contested at different levels. The paper then concludes with a discussion of the effectiveness of the policies and structures in place at the national level to address gendered inequalities.

 

2. Gendered inequalities in Ethiopia  

 

Ethiopia is recognised as one of the most impoverished countries in the world characterised by persistent levels of extreme poverty and inequality. It has a per capita GNP of approximately US$100 (ADB, 2004) and ranks 170 out of a total of 177 countries in the Human Development Index (HDI, 2005). Agriculture is a main source of livelihood for the majority of the Ethiopian population who are poor and rural (World Bank, 2005).  Ethiopian society is diverse and complex; it is deeply hierarchical and characterised by a distinct gendered division of labour crosscut by ethnicity, religion, class and age.

 

Women comprise approximately 65% of the informal sector and provide 60% of the total agricultural labour (ADB, 2004). In addition, they perform a key role in maintaining and managing the household through their responsibilities for reproductive activities such as food preparation, health and hygiene and childcare. Women perform longer working days than men, yet lack access to modern technologies that would alleviate their tasks since these tend to be appropriated by men.  Much of their time is spent in walking long distances to fetch water, collect fuelwood, and preparing food through labour intensive tasks (e.g. milling flour). In spite of this key role in the Ethiopian economy and household welfare, they experience persistent inequality and discrimination. Women are culturally regarded as inferior to men and are marginalised by hazy legal rights and male biased institutions and markets which ignore the needs of women farmers (Tadesse, 2001).

 

Although there has been a national improvement in some welfare indicators such as education where national primary gross enrolment in grades 1-4 increased from 30% in 1994-1995 to 83% in 2000-2001 (World Bank, 2005), women consistently perform poorly in measures of education, mortality rates, health and other key social and economic indicators. For example, Ethiopia’s performance on the gender gap in primary school enrolment and adult literacy and the representation of women in government is consistently below the average of the whole of sub-Saharan Africa (World Bank, 2002). There is also evidence suggesting that more women are likely to be infected with HIV/AIDS and less likely to have knowledge concerning the disease and awareness of how to prevent it.  Access to family planning and choice of methods also remains limited and constrained by husband’s approval. Women also lack decision making power in the household and community as well as access and control of productive resources. In addition, women are subject to multiple ‘harmful traditional practices’ such as female genital mutilation, early marriage and rape (ADB, 2004; World Bank, 2005; NCTPE, 2003). The explanation of these persistent inequalities can be traced back to a deeply patriarchal society that draws on tradition, culture, law and custom to legitimise the widespread continuation of the subordination of women. How has the government of Ethiopia (and donors) dealt with this? What policies has it implemented to challenge the foundations of unequal power relations?

 

3. Gender mainstreaming in Ethiopia

 

Gender mainstreaming is used to describe both the process of institutionalising gender equality within the development context (Goetz, 1997) and the strategy of assessing the effects of policies on women and men through gender planning (Saunders, 2002; Bell et al, 2002; UNDP, 2000, 2003; Byrne et al, 1996; UN, 1997).

Under the influence of Gender and Development (GAD)[4] discourses, gender mainstreaming is interpreted as seeking to produce “transformatory processes and practices that will concern, engage and benefit women and men equally by systematically integrating explicit attention to issues of sex and gender into all aspects of an organisation’s work” (Woodford-Berger, 2004:66). This redirected attention away from increasing women’s participation towards looking for ways to transform the development agenda itself.

 

There is a general consensus that gender equality concerns can be mainstreamed in two interdependent ways: integrationist and transformative or agenda setting (Jahan, 1995; Mukhopadhyay, 2004; Kanji, 2003). An ‘integrationist’ approach focuses on the supply side (or technical side) of GM by ensuring that “gender equality concerns are integrated in the analysis of problems faced by the particular sector” which is then used to inform policy and practice using targets that are measured using a range of sophisticated monitoring and evaluation tools, frameworks and checklists (Mukhopadhyay, 2004: 96). The latter stem from the rigorous academic and theoretical background of GAD advocates who intended to strengthen technical[5] capacity to enforce the incorporation of gender equality concerns into all aspects of development policy and practice (Mukhopadhyay, 2004).

 

In contrast, the ‘transformative’ approach focuses on changing the development agenda itself by creating the demand for change with the introduction of women’s concerns in relation to their position[6]. This involves a deeper understanding of the context in which this change is to be enforced such as state-society relationships, political society characteristics and the influence of international development and financial institutions on policy making and practice (Mukhopadhyay, 2004). This moves GM beyond a technical exercise to include a political process whereby agendas, institutions and organisations are changed; analysis shifts towards relations of power and inequality rather than gender roles; and intervention is broadened beyond projects to include programmes, partnerships, policy processes and agencies themselves (Kanji, 2003). 

 

In response to the glaring inequalities highlighted in the previous section, the government of Ethiopia has made concentrated efforts to ‘mainstream gender’ and promote the empowerment of women after the fall of the Derg (1991) through various policies, initiatives and programmes.  Based on a review of secondary literature and relevant policy documents, the evidence suggests that Ethiopia has pursued an integrationist approach to gender mainstreaming which has not quite taken off. The following discussion seeks to summarise key policies and initiatives aimed at mainstreaming gender and highlight areas of constraint.

 

Before more recent policy documents are discussed, it is important to note the Derg regime (1974-1991) did make efforts to preach the goals of women’s emancipation through its socialist ideologies. These messages were spread through Revolutionary Ethiopian Women’s Associations (REWA) whose overarching goal was to raise the economic and political position of women (Pankhurst, 1992).  The official propaganda of REWA stated that its aims were to:

 

·         To propagate to women the theory of scientific socialism with a view to raising their political consciousness and cultural standards

·         To prepare women to occupy their appropriate position in society and to participate actively in productive social activities

·         To make every effort to ensure that the rights of women as mothers are recognised and they, as well as their children, are well cared for

·         To liberate women from political, economic and social dependence and prepare them to join hands with their class allies and fully participate in the struggle to build socialist Ethiopia (Ethiopia, Revolutionary Ethiopia Women’s Association, 1982: 25-6, cited in Pankhurst (1992))

 

Although REWA did have a significant role in raising women’s awareness, it had limited success in implementing these rights in part because of the authoritarian nature of the political regime (Michael, 2000). Pankhurst (1992) notes how their effectiveness was limited to the level of rhetoric since they lacked real power and were plagued by administrative inefficiency.  Nevertheless, it did set the scene for the new government to take on the task of gender mainstreaming in the context of growing emphasis in the international arena of the importance of integrating gender into the development agenda.

 

The Ethiopian National Policy on Women (1993) was an important first step towards mainstreaming gender. The main goal of the policy was to facilitate the conditions for equality between women and men. To this end, it demanded intervention to:

 

·         Mainstream women into existing laws, regulations and customary practices to allow women to participate in decision making structures

·         Coordinate and incorporate women’s issue in all government programmes and policies at all levels

·         Change discriminatory attitudes towards women and girls

·         Promote research and awareness in all areas concerning women’s development and equity.

 

Of far greater importance was how it stimulated plans for implementation through the creation of gender mainstreaming infrastructure to institutionalise the political and socio-economic rights of women.  These took the form of a Women’s Affair Office (WAO) under the Prime Minister's Office; Women Affairs Departments (WADs) located in each line ministry; Regional Women’s Bureaus (RWBs) at the regional administration level; and a women’s coordination and desk officer at the Zonal and Woreda level. The overall responsibility of the WAO is to coordinate and implement all aspects of the National Policy for Women at national, regional and sectoral level. In other words, it is the main body responsible for mainstreaming gender throughout the country. The WADs are responsible for addressing gendered inequities in their particular sector and are supposed to liaise closely with the WAO.  In comparison, the RWBs, who are accountable to their respective Regional Administrative Council, are responsible for gender mainstreaming at the regional level. The zonal and woreda level desk officers work more closely at the community level to identify and address gender needs. They also play an active role in implementing various gender related programmes and projects. The detailed responsibilities of these machineries are detailed in Table 1 below (ADB, 2004).

 

Table 1: Responsibility of gender machineries

 

Women’s Affairs Office (WAOs)

 

Women’s Affairs Departments (WADs)

Regional Women’s Bureau (RWBs)

Location:

Prime Minister’s Office

Location:

16 Line Ministries

Accountable to line ministry and report to WAO.

Location:

Accountable to Regional Administrative Council

Coordination, facilitation and monitoring of women’s affairs activities at the national level

Create favourable conditions for effective gender mainstreaming and implementation of gender sensitive activities in their particular sector

Provide gender mainstreaming guidance to all Regional, Zonal and Woreda level programmes and project interventions

Initiate proposals for developing gender sensitive policies as well as reviewing existing ones to ensure that gender is mainstreamed

Monitor progress and report to the WAO periodically on challenges and constraints

Identify gender mainstreaming and equity needs at all levels in region

To collect relevant data and information nationally and sectorally in order to disseminate them as well as to undertake studies on pertinent gender and women’s issues

Ensure that gender is mainstreamed in all projects and programmes within the authority of their respective sector and ministry

Develop respective gender mainstreaming strategies for the different technical and sectoral bureaus in the region

To organise workshops, conferences and symposiums at the national level to promote and raise awareness on women and gender issues in the Country

 

Initiate, undertake and disseminate critical research and studies related to gender and women’s development issue within the regional context

To enable a conducive environment which will promote women’s empowerment and equal participation

 

To initiate and undertake gender mainstreaming and awareness raising session at all  levels of region and across all sectors

 

 

Create favourable conditions for implementation of NPW and monitor various activities. Identify areas of concern base on needs and priorities of each region and plan to achieve those goals

 

Source: ADB (2004)

 

The efforts of the WAO are also supported by a number of bilateral and multilateral agencies that have played a greater role more recently in sharpening the commitments of the National Policy for Women. These include the UN Interagency Working Group on Gender consisting of gender focal points from all agencies and the Group for the Advancement of Women (GAW) (later renamed the Donor Group on Gender Equality). The United Nations Development Assistance Framework (UNDAF) is also an important programme aimed at facilitating development more generally, but gives specific emphasis to gender (Sida, 2003). Membership of GAW consists primarily of bilateral and UN agencies together with some representatives from various NGOs, EU, OAU, ILO and IOM. It has played a key role in moving forward from the Beijing platform of action commitments for the follow-up Beijing +5 conference in 2000. These efforts are also supported by various organisations and women’s associations such as the Ethiopian Women Lawyers Association (EWLA), the Network of Ethiopian Women’s Associations (NEWA) and the Centre for Research Training and Information for Women in Development (CERTWID) who are all committed to assisting women to empower themselves and benefit from the development process.

 

In addition to the National Policy for Women, the Ethiopian government’s commitment to gender mainstreaming is explicit within the Ethiopian Constitution (1995) which is the ‘supreme law of the land’. The Constitution enshrines a commitment towards the equal rights of women with men and recommends affirmative action to address past inequalities.  Article 35 also asserts that women have the right to land, equal rights in marriage, and the right to be protected from harmful traditional practices (Demessie et al, 2004). It also reaffirms Ethiopia’s commitment to various international conventions focused on gender equality such as the Convention on the Elimination of All forms of Discrimination Against Women (CEDAW) (1981) and the Beijing Platform of Action (1995) in addition to broader human rights conventions[7].

 

The Constitution has provided an important foundation for further revisions of discriminatory laws. The new Family Law has been influential in amending discriminatory provision relating to marriage, divorce and the upbringing of children (NEWA, 2004; NEWA & EWLA, 2003). This process has been greatly assisted by the Ethiopian Women Lawyer’s Association (EWLA)[8].  The Family Law has succeeded in raising the legal age of marriage for girls from 15 to 18; asserted that the contract of marriage is between spouses and not between the family; maintained that any matrimonial property should be shared and that the husband must consult with his wife when making decisions concerning property. This has been followed by further revisions of the 1957 Penal Code which came into operation in 2005. This dealt specifically with addressing issues related to rape, abduction, domestic violence, female genital mutilation and abortion. One of its most significant changes is with regards to abduction. Now the act of marriage following abduction does not cancel the crime and the act of rape following abduction is still punished. In the past it was not considered a crime if the couple were married afterwards (EWLA, 2004). It has also made female genital mutilation illegal and introduced penalties for perpetrators. However, it has been criticised for failing to include provisions on sexual harassment in schools and workplaces; restraining orders against perpetrators wishing to pursue revenge and there are no provisions of psychological violence.

 

In parallel to these key laws and legislations, there have been a number of other key policy documents aimed at creating the conditions for sustainable development which have included some mention on provisions for gender equality. These include the Agricultural Development Led Industrialisation (ADLI) programme, the National Population Policy, the Education and Training Policy, Health Policy, Development Social Welfare Policy, Environmental Policy, Cultural Policy and Policy on Natural Resources and Environment. It is beyond the remit of this paper to discuss this in detail and more information can be found in a report compiled by Haregewoin Cherinet and Emebet Mulugeta commissioned for Sida (2003).

 

The extent to which goals of gender equality are integrated within the Poverty Reduction Strategy Papers (PRSP) is also a good indicator of commitment towards gender mainstreaming as they are the overall national strategy for reducing poverty. Ethiopia’s first PRSP called the Sustainable Development and Poverty Reduction Programme (SDPRP) (2002-2005) recognised gender as a crosscutting issue and made a clear commitment to gender equality in the context of the overall development strategy (Moritz & Musisi, 2004) as illustrated below:

 

“Core dimensions of poverty (opportunity, capability, security/risk, and dis/empowerment) differ along gender lines, and function to heighten the vulnerability of women. For these reasons, the inclusion of gender in any effort to alleviate poverty is non-negotiable” (SDPRP, 2002: 122)

 

In spite of these key policies and laws in place, there are loopholes. With regards to the SDPRP, there is no indicator for promoting the empowerment of women and gender equality within its overall matrix of indicators and targets for monitoring progress. Moreover, the performance indicators that do exist are gender neutral. A major constraint for the Constitution is that it recognises customary and religious laws relating to personal relations; these include marriage, divorce, property ownership child custody and inheritance adoption (Birke et al, 2002). This means that some of the progress made with the other laws are made void. Another key constraint concerns the separation of jurisdiction and authority between the federal and the regional state contained within the Constitution.  Regional states have the power to enact family codes whereas the Federal state retains authority of all other areas of civil law. Constraints arise when regional Family Laws may include elements of customary practices that indirectly discriminate against women. As a result, many regions still practice the old Family Law where the husband is the head of the family and has the right to make all major decisions regarding property.

 

Indeed, many of the reports consulted for this summary stressed that having the appropriate laws and policies in place were no clear guarantee that their recommendations would be implemented.  Indeed, there is still widespread evidence suggesting that FGM continues and that early marriage is prevalent in many areas (World Bank, 2005).  A more fundamental constraint is the lack of awareness of women’s rights by women themselves. A recent study by EWLA on violence against women in Addis Ababa (2003-2004) notes that many forms of violence such as beating goes unreported because the women didn’t know that they could report; were ashamed or afraid to report; or felt that reporting would result in additional violence.  A recent World Bank report (2005) also notes that different women have different abilities to report violence. For example, consultations with EWLA revealed the difficulties of bringing a rape case to court if the victim’s virginity is in doubt. As a result, EWLA is campaigning that non-virgins should be given equal consideration before the law as they are just as vulnerable to rape.

 

On the whole, the majority of the reports consulted for this summary together with findings in the WeD research concur that customary practices and the ideas of conforming to cultural norms play a key role in reinforcing and maintaining gender inequalities and are therefore a key obstacle to the implementation of gender sensitive policies. They also conclude that although there is much rhetoric on women’s rights and gender equality within policy, laws and regulations, little is done in practice. Part of the explanation for the failure of translation of policy into practice is due to a lack of awareness amongst women themselves and the desire to conform to cultural norms. These difficulties are further compounded by the very infrastructure delegated with the responsibility of gender mainstreaming; and it is this constraint, when addressed, which has the potential to transform gender relations for the better.

 

The Women Affairs Office has the primary responsibility for creating a conducive environment for gender mainstreaming and monitoring progress. However, various reports indicated that the effectiveness of the women bureaus at the federal and regional levels were hindered by a number of factors (Sida, 2003; ADB, 2004; NEWA, 2005; World Bank/WAO, 1998).  These include a lack of clear mandate and authority; poor institutional capacity and lack of capital budget (World Bank/ WAO, 1998).  By far the greatest constraint concerns the issue of capacity, particularly in relation to staffing. Many of the delegated staff perform dual roles: they work for their particular line bureau and the women’s bureau. Often responsibilities in the former outweigh the latter. This is accompanied by lack of clear mandate and responsibility which is particularly acute for the WADs and is linked to an overall lack of accountability for gender issues. At the federal and regional level, Women Affairs machineries have a limited role in key decision making in policy and practice and are not even a member of the cabinet[9]. They therefore lack authority to enforce changes towards more gender sensitive intervention on other line Ministries in spite of the WADs in place.  This explains why there are few national sector programmes including gender and few ministries that prepare gender mainstreaming guidelines Decentralisation has also served to marginalise gender issues by complicating the issue of accountability between Federal and Regional levels. Lack of staffing also limits the extent to which staff can liaise with other stakeholders engaged in gender mainstreaming as well as their interaction with communities on the ground. For example, there is little interaction with local based institutions such as iddirs (burial associations) to address these issues.

 

Another problem highlighted by a recent study conducted in preparation for National Action Plan on Gender was a lack of awareness of the key policies and laws affecting gender rights amongst the woreda officers and even the regional women bureaus to certain extent. This is the outcome of a lack of regular and context specific gender training of staff at all levels which inhibits the overall success of gender mainstreaming initiatives. A lack of adequate financial resources committed towards gender mainstreaming is another key constraint. This inhibits the ability to collect gender disaggregated data which is essential for monitoring progress. At the local level, it restricts the recruitment of staff and places transport constraints on woreda officers carrying out delegated responsibilities. Often, their activities are restricted to ‘organising women’ rather than conducting gender analysis and educating local women on their rights. This is compounded by an overall sense of demoralisation with gender mainstreaming related to how the WAO has been marginalised resulting in high staff turnover (NEWA, 2004).  These problems are not uncommon and characterise many gender mainstreaming structures across the world. Indeed, Longwe (1997) characterises the tendency of gender policies to disappear within bureaucracy as the evaporation of gender policy within the ‘patriarchal cooking pot’.  

 

The roots of these difficulties can be traced to the inherently political nature of the transformation that gender mainstreaming aims to achieve.  As mentioned earlier, gender mainstreaming is a political project which requires a fundamental shift in the way that development intervention is implemented and this is why it is ultimately resisted. This resistance is embedded in the continuation of a policy making infrastructure which remains male-biased and patriarchal. There is a fear of the change in power balance between women and men which partly has its roots in people’s cultural attitudes to gender roles and relations.  It is also partly a function of a lack of awareness, and more importantly an understanding of what gender mainstreaming entails.  It is this bias, and the assumption that gender is a ‘woman’s issue’ that underpins the weak financial and human resource capacity of gender mainstreaming structures. It comes as no surprise that there is high staff turnover within these gender machineries (NEWA, 2004). Ultimately, this weak capacity perpetuates a vicious cycle because it hinders the authority of these structures to assert mechanisms of accountability. The result is that gender mainstreaming can be perceived as elusive and nebulous without a clear agenda for transformative action, thus producing diverse strategies to mainstream based on patchy understandings of what these processes are meant to achieve (Subrahmanian, 2004; Woodford-Berger, 2004 ).

 

Thus far, the evidence suggests that there have been consistent efforts to create the infrastructure and policies to commence an ‘integrationist’ approach to gender mainstreaming. However, to date, it appears that such efforts have been impeded by a number of reasons mainly due to weak capacity amongst the gender mainstreaming infrastructure. Nevertheless, some progress is being made to improve the effectiveness of an ‘integrationist’ approach.  Although the former SDPRP has been criticised for failing to systematically mainstream gender throughout the whole strategy and not providing adequate gender disaggregated data on poverty indicators, it has had some positive outcomes. In particular, it has been instrumental in leading to improved dialogue between the Government of Ethiopia and donors resulting in various new structures to facilitate the implementation of the strategy. Of particular reference to gender mainstreaming was the creation of Joint Group on Gender Equality (JGGE) comprising government officials and donors and a Donor Group on Gender Equality (DGGE)[10]. The key function of these groups is to improve, coordinate and harmonise gender mainstreaming throughout government policies, programmes and projects, particularly the SDPRP (DGGE ToR, 2004). This has become particularly important towards developing a strategy to meet the Millennium Development Goals. The SDPRP together with the aforementioned groups has also stimulated a significant number of new initiatives aimed at translating the National Policy for Women and the various commitments to gender mainstreaming in other policies (e.g. SDPRP) into practice. These initiatives include the National Action Plan on Gender and Gender Budget Analysis, together with better gender-disaggregated reporting.  These initiatives were underway towards the end of the SDPRP and were reincorporated into the second PRSP called the Plan for Accelerated and Sustained Development to End Poverty (PASDEP) which will run for five years from mid 2005 to mid 2010.

 

The National Action Plan on Gender (NAP) was formulated specifically to put the commitments to gender mainstreaming scattered across the various policies into action (PASDEP, 2005). Specific objectives include:

 

  1. To mainstream and articulate gender concerns on broad policy processes including the PASDEP, MDGs and budget process
  2. To strengthen gender analysis and overall gender sensitively of the PASDEP monitoring and evaluation system (including the core indicator set, data collection methods)
  3. To strengthen the Women Affairs Office (WAO) institutional structures and functioning, and identify entry points/activities for enhancing gender mainstreaming across government policies and programs

 

Progress thus far is that a draft NAP has been completed and plans for implementation are underway.  With regards to the Gender Budget Analysis, this is a response to the recognition that policy implementation can be approved through Government Budget. The aim of the initiative is to analyse pubic expenditure from a gender perspective, particularly how it has affected women and men as users of public services. Its ultimate goal is to allow government to use a budgetary process to address gender inequality.

 

These steps represent important progress towards an integrationist approach to gender mainstreaming which will kick-start the more transformative agenda of gender mainstreaming. However, time will be the ultimate test.

 

4. Gendered inequalities on the ground

 

How do these initiatives to mainstream gender and overcome gender inequalities translate at the grassroots at the level of household and individual?  How does policy deal with the factors that maintain unequal power relations? The following discussion draws on the analysis of power and poverty by Bevan, Pankhurst and Holland (2005) in conjunction with preliminary findings from WeD Ethiopia’s research in four rural communities[11] (DEEP[12] sites). The research revealed that gender crosscuts with class, age, ethnicity and religion in different ways to produce imbalanced power relations that disproportionately affect women.  The evidence suggests that women are subject to ‘controlling power’ structures which describe the structural relationships of exploitation, exclusion, domination and destruction found within households and communities (Bevan et al, 2005). They are also negatively affected by ‘dispositional power’ which comprises the structuring of symbolic resources or capital, including knowledge and information, political ideologies, religious beliefs and customary values.  These two types of power in turn greatly constrain a woman’s individual agentic power which describes the capacity of individuals to make potentially life-transforming decisions and to act on them[13]. The following discussion will now explore these dimensions of power over four facets of women’s inequalities. These are the gendered division of labour; the pregnancy/infancy/birth cycle; marriage and violence against women’s bodies.

 

4.1: Gendered division of labour

 

The gendered division of labour is a key mechanism of reinforcing unequal relations between women and men. It legitimises the heavy workloads of women which are underpinned by relations of exploitation and domination by men (i.e. controlling power structure). The following quote reveals how ideas of ‘appropriate’ gendered behaviour are reinforced through gendered roles. Men are perceived as the decision makers, the ones who control assets, provide security for the household, participate in community affairs and gain respect for their responsibilities. In contrast, women are expected to perform a ‘double shift day’. They are seen as being primarily responsible for traditional female reproductive and domestic roles such as food preparation and childcare. A wife is responsible for fulfilling her husband’s every want. She should be demure and respectful of her culture and allow her husband to make the major decisions.

 

Wond (male) Head of the household manage and control household resources; main rules are engaging in farming and giving social and psychological security to household members; main rights are respect from hh members, make contractual agreements, not doing house activities which are the responsibility of women.

 

Set (female): main responsibilities are child-caring, food preparation etc, fulfil whatever husband needs; main rules are not to participate in local meetings which includes males only (female response); treatment of husband, to respect the culture, not to be talkative (male response) . Main rights – second head of household, giving advice to her family (female response) preparation and distribution of food to family members, respect from children. (Yetmen: WIDE2)

 

These norms are reinforced through community gatherings where women are expected to be absent or ‘quiet’ at community meetings to allow their husband’s to take an active role. These rules can be bent when husbands are absent, or when women are single and/or are part of an iddir (customary burial associations) or mahiber (association dedicated to a saint[14]).

 

Women are expected to be confined on their houses and do household activities. They will only attend outside events when their husbands are not around or they are single and have their own iddir or mahiber. Women have equal right to own properties with men. They can only sell or exchange something with the permission of their husbands. (Yetmen)

 

Any resistance to this set formula is often perceived as disrespecting culture and religious upbringing, and is consequently severely frowned upon (particularly from the older generations). However, poorer women seeking a means of survival are faced with no choice but to take on labour which is seen as ‘shameful’ for women because it is not regarded as traditional female activities as illustrated below.

 

 

Adult females do not go to other rural areas for work. But there are about 2/3 women, who are very poor, who work at Hamus Gebeya (they prepare Katikala  (alcoholic distilled drink) four/six days a month). One of the respondent states that the community consider it as shameful to go to rural area for work for adult female. (Turufe Kecheme)

 

 

Young girls and boys are socialised into these roles and unequal gender relations from and early age through the careful guidance of their parents and religious upbringing. Although there is a slight difference in gender roles amongst Christians and Muslims (i.e. in that the Muslim girls are more restricted in what activities they can take on), there appears to be agreement that a young girl’s responsibility is to become proficient in carrying out ‘female activities’ within the sphere of the household.


 

Both girls and boys should start working in the house at 3, as messengers from about 3 years of age and began working out side at 4. Boys should begin to go out in the fields with fathers or shepherds in order to prepare them for work (at 4) and at 6 they should start herding in assistance with elders at the age of 6 and independently at 7. Then hoeing, clearing crop fields, weeding, levelling threshing floor as they grow 10. At 6 girls should be able to roast and boil coffee, collect fire wood, fetch water, clean house and begin to make Injera, Wat. (Dinki: Young Lives 1M  Poor  man)

 

Girls should clean houses, make the fire, roast/boil coffee, collect fire wood, fetch water and Christian girls should also herd animals and weed. Both Christian and Muslim girls should completely manage to work all female activities at the age of 10.  (Dinki)

 

Increasingly, there is evidence suggesting that both women and men are beginning to challenge these gender roles in nuanced ways. For example, in our protocol data from Korodegaga we have a case study of one man with 3 wives who supports his 15 year old daughter in her education against her mother who wants her to get married. In Dinki, we have another case where the male household was supportive of his children going to school. However, because of labour shortage he had to retain his eldest son and daughter to assist with household and agricultural tasks whilst the younger children (including girls) went to school. This illustrates the complexity of gender relations which are crosscut by the circumstances of the household.

 

More subtle variations in gender roles are witnessed across religious denomination (whether you are Muslim or Christian), ethnicity and class/wealth. The following quote from the Young lives protocol in Dinki (a predominantly Muslim site) reveals how Christian Amhara girls are taught to perform agricultural fieldwork skills in addition to traditional female domestic tasks within the household.  They also appear to be more mobile then Muslim girls who tend to be restricted in the places they can visit (e.g. markets).  Religion crosscut by ethnicity also appears to affect parent’s attitudes to education. Muslim Argoba parents tend to view education as having a bad influence over their child’s “survival” whilst Amhara Christian parents see education as a way of increasing economic opportunities as illustrated below. However, it is important to note that the Argoba tend to favour Muslim education which is different from secular education. So this moves the question beyond the issue of having education or not, towards the type of education.

 

Muslim and Christian parents have different views and practices as far as genderization of skill learning Muslim Argoba believe that it is a religious duty that girls should learn only home making work skills and remain in the house. But the Amhara Christians believe that girls must learn important fieldwork skills too. Thus both category parents revealed that unlike Muslims, Amhara girls equally begin to be engaged in herding as boys.

 

The Muslims never allow girls to go out to markets or other events where as rich/middle Christians could assign some one to protect their girls if they go out to festivals or markets.

The two sub groups seem to maintain contrary values and attitudes to wards religious and formal education from gender perspective and goal orientation. Argoba Muslim parents highly value religious education as an appropriate means of socializing children and it enables them to acquire the required behaviour as their survival strategies. On the contrary they believed formal education will have effect against their children's present or future survival in the community. The Amhara/ Christians rather consider formal education as positive survival/strategy against logically and nationally defined economic resource problems. Although Amhara parents seem to agree with Muslims on misdirecting effects of formal education in terms of behaviour, they value it in terms of economic survival strategy. (Dinki: Young Lives 1M)


 

For young girls, this socialisation process is an important means of preparing them in ‘housewifely duties’ to make them good wives and mothers.

 

An adolescent girl who does not fulfil all housewifely duties may not get married and will be given the nickname geltu (lazy). (Yetmen)

 

Being a good wife and producing offspring (particularly boys) is crucial for affirming female identity. Women who produce sons are rewarded with praise and respect from husbands and other women since males are seen as being able to “defend themselves and their family better from any danger” whilst females are “easy victims for enemies”.  For this reason, male infants receive more care and nourishment than female children.  Only when the girls become of marriageable age, do they receive more attention and care from their mothers as it is important for her to find a husband who will provide an important source of support for the parents in their old age. Alternatively, when a mother has many children, she needs more female children to help her take care of the domestic household tasks.

 

There is a belief, particularly among Orthodox Christians, that it is better to give birth to male than female children. They believe that males can defend themselves and their family better from any danger while females are easy victims for enemies. A husband feels happiness when his wife gives birth to a son and the wife also feels proud. The husband may kill a sheep or a goat for the wife who gave birth to a son. When a son is born, women who gather in the house of the woman giving birth make a thin loud clamour called ililta seven times, but only three times if the child is female'. (Turufe Kecheme: WIDE1 1995)

 

Women and girls who fail to fulfil these gender norms, particularly those who do not produce children, are generally seen as failures and treated with disdain as illustrated below.

 

Even if the husband may be infertile it is the woman who is blamed: she is labelled as beklo (mule) and considered to be cursed. (Yetmen)

 

However, we do have a case in Dinki where the couple have not had children but seem happy and the husband treats his wife well; thus raising the importance of interrogating generalisations and recognising diversity.

 

 

4.2: The pregnancy/birth/infancy cycle

 

A key factor compounding women’s inequality that deserves being considered separately as a distinct dimension of the gendered division of labour in its own right is the biological role that women perform in society. If we look at this in power terms, it is an important means in which women’s bodies are dominated and exploited. A key finding of the research was that women’s childbearing and childrearing responsibilities and related needs were rarely recognised in prevailing discourses on gender relations. However, research by WeD Ethiopia in 20 rural WIDE[15] sites in 2004 summarised by Bevan (2004)[16] below revealed that women experienced difficulties in relation to infertility, pregnancy outside marriage, abortions, spontaneous miscarriages, pregnancy, delivery and infancy. These were further exacerbated by seasonal changes in food, cash and water availability, temperature, disease outbreaks and the timing of women’s agricultural labour.

 

Infertility was a key problem for rural women in Ethiopia as failure to become a mother challenges the very foundations of female adult identity and status. Not only does it cause profound personal unhappiness, it can also result in divorce and stigmatisation by the rest of the community. In the long run, it also means that there is less labour for the household and insecurity in old age.  Paradoxically, when infertility is due to the male, there is little criticism and often the woman is blamed.

 

Much female infertility is caused by abortions and miscarriages often associated with child pregnancies (due to early customary marriages) and circumcision. Pregnancies outside of marriage are regarded as taboo, and many resort to abortion using unsafe methods such as malaria tablets or drinking bleach to avoid stigmatisation. Women who experience spontaneous miscarriages are blamed. Lack of the appropriate care (or ability to afford it) can lead to further complications and immense pain that can result in death or infertility.

 

The experience of pregnancy and delivery can also produce much suffering for women. Faced with the responsibility for other domestic tasks within the household, women struggle to reach a balance whilst dealing with the effects of pregnancy such as sickness, fatigue and pain. This is particularly difficult during times when agricultural work in the fields is crucial (e.g. weeding, sowing and harvesting). During times of drought, hunger can cause negative effects to the unborn child as they fail to attain adequate nutrition to develop. Female circumcision can produce considerable problems for the delivery of the child that can result in death for both the mother and infant.

 

Bevan notes that “a major maternal worry” is the health of the baby. When an infant falls ill and traditional methods of healing fail, not all can afford the costs of formal health facilities. Young children are also more prone to droughts and disease epidemics. Because young boys are favoured over girls, they tend to be fed more, making young girls particularly vulnerable.

 

In summary, Bevan outlines three key issues that create hardship for women: food and diet for mother and baby; work and time-use in relation to women’s’ multiple roles within the household; and maternal and baby illness and health care. The food and diet of both the mother and baby is crucial throughout the pregnancy infancy cycle. Their propensity to find enough food to survive (to avoid starvation) and find the right types of food to flourish (nutrition) is dependent on their physical location and relative wealth.  Women’s’ multiple roles in the production and maintenance of the household together with their biological role as mother means that they have to find a delicate balance within their ‘double shift day’. Lastly, the pregnancy/infancy cycle is a particularly vulnerable time for mothers and babies that requires reliable preventive health measures and sufficient health care to mitigate pregnancy related illnesses. However much of this care is of poor quality, unaffordable or non existent.

 

The extent of acute and chronic suffering experience by mothers and babies during the pregnancy/infancy cycle is summarised in the table overleaf:


 

Table 2: Acute and chronic suffering experienced by mothers and baby

 

Mothers

Babies

Infertility

Unwanted pregnancies

Too many pregnancies

Botched abortions

Miscarriages

Difficult pregnancies

Illnesses during pregnancy

Starvation and nutritional failures during pregnancy

Overwork during pregnancy

Absence of, poor, or costly health facilities during pregnancy

Lack of care and support during pregnancy

Difficult deliveries

Absence of, poor or costly health facilities during pregnancy

Lack of care and support during pregnancy

Difficult deliveries

Absence of, poor, or costly health facilities during delivery

Post delivery complications

Absence, of, poor or costly facilities for new mother and baby

Overwork once baby is born

Starvation or nutritional failures whilst breastfeeding

Being responsible for the physical welfare of the children, especially in times of famine

Worry about feeding the family

Worry about the health of their baby

Nutritional failures while in the womb

Problematic birth experiences

Starvation or nutritional failures in the first trimester

Local diseases and illnesses

Neglect which may be result of mothers having to make ends meet in hunger periods or because the child is female (i.e. due to cultural preference for boys)

 

Source: Bevan (2004)

 

 

4.3: Marriage

 

The main purpose of marriage is to strengthen or establish new social capital networks by joining families, kin groups of clans. These networks become important sources of support at times of stress (famine or death of family) or labour during the key agricultural periods. Being married is important for a woman’s adult status and identity.  Being unmarried risks being labelled a “failure”. However, the mechanisms through which couples are ‘married’, the actual experience of marriage, and the impact of ending marriage (through widowhood or divorce) play a profound role in entrenching and legitimising gendered inequalities.

 

The research revealed different marriage rules amongst the sites according to ethnicity and religion.  Yet, all sites practice virilocal marriage residence rules where women are expected to move away form their parental village to that of the groom and often into or next to his parent’s home.  The implication of this move is that a woman is often socially and geographically isolated form kin and friends. For the Oromo of Korodegaga (All Muslims) and Turufe Kecheme (mostly Muslims) and the Argobas of Dinki (all Muslims) bridewealth is often given to the wife’s family from the husband’s family.  This can include cash or material items such as clothes or livestock to help support the birth of new children.  Often there are strict rules about who one is able to marry and marriages are frequently arranged by parents. Within the research, there were many cases of young girls reporting that they would return home to find parents had arranged a marriage against their will. On the whole, women have little say in the matter as illustrated below.

 

Marriage within a clan is forbidden; it is not an individual affair but the concern of the clan – it unites two clans. Forms of marriage include marriage by purchase, marriage by exchange, abduction, by consent of partners without parental agreement, marrying dead sister's husband, marrying dead husband's brother, marrying a divorcee. (Korodegaga: WIDE 1)  

 

More violent forms of marriage through abduction (involving kidnapping and rape) were frequent. This proved to be a strategy used by poorer men to either avoid or reduce the large bridewealth payments. This form of marriage caused much distress for young girls who experienced the terror of rape. Divorce was not an option because of the fear of stigmatisation from the community as illustrated below

 

During my marriage, I was very sad because I married through abduction. Since divorce was considered as devotion from accepted norms, I chose to stay with my husband for some years. (Turufe Kecheme)

 

Polygyny is also prevalent amongst the Muslim Oromos and can have negative consequences for wives and children where tensions between wives can lead to competition for resources within households and mistreatment of children. The practice of having multiple partners has also contributed to the spread of sexual diseases which have disproportionately affected women.  This is compounded by adultery amongst men.

 

The inheritance of a widow by a dead husband’s brother or through the replacement of a dead wife by her sister is still practiced amongst Muslim Oromos. This arrangement can create similar constraints to women as those entrenched in polygynous relationships as illustrated below.

 

I was 14 years old married. This was the worst thing in my life because I married my sister’s husband when she died. This was a big impact in my life. Another impact in my life was the death of my husband. In 1993, I was married to a second husband. I am pregnant with the second child for my second husband. But he married another woman, so he didn’t help me. (Korodegaga: ALF widow 24)

 

The Amhara are the dominant ethnic group in Yetmen and a large proportion in Dinki but are a minority in Turufe Kecheme and Korodegaga. Although Amhara rules differ, they still serve to embed the view that the woman is the possession of the man and that her purpose is to run the household and produce children. It is common amongst the Amhara to marry early which in turn results in early pregnancy which create difficulties of young girls as their bodies are not fully developed to handle pregnancy.

 

For Amharas marriage is a contractual agreement involving the pooling of labour and property to establish a new household and raise children. Three recognised types of marriage are 'equal partner', communion marriage (indissoluble), pay marriage (monthly salary). There is also living together. (Yetmen)

 

These different marriage types are accompanied by different rules regarding widowhood and divorce. Fafchamps and Quisumbing (2002) give a detailed account of how the control and ownership of assets following death or divorce varies according to ethnicity, the degree to which assets were brought into the marriage (from the wife’s side), and religion; thus illustrating the complexity of gender relations under different circumstances[17]. They also note that government’s policy on land and conversion to non-Orthodox Christian faith is having some impact on the allocation of assets after death or divorce, but recognise that this needs more investigation.

WeD research revealed that divorce from marriage through bridewealth (common amongst the Oromo Muslims) usually results in the divorced wife losing all her rights to property as illustrated below.

 

A divorced wife has no right to take any property which she and her husband got after marriage; she can only take her clothes and gifts she received from her parents during the marriage which may include furniture and cows. (Turufe Kecheme)

 

In Muslim marriage rule there is a statement that is signed by the couples in case divorce happened. The wife should wait six months if her husband permit to marry with other man. And there is also rule how they divided their property or land or their livestock. The wife couldn’t get anything from the property or land except the endowments that she get from her parent; otherwise she returned back with thirty birr only. But there are differences in divorcing by now Muslim females get 70 and above money during divorcing. (Dinki: WIDE2)

 

Although women have greater claims to property amongst the Amhara (see below), in reality their ability to secure this right is constrained. This is particularly the case for women with no source of independent income (Pankhurst, 2006). As a result, many may decide to withhold their right and migrate back to their home village.

 

During divorce children may live with their father or mother depending on the parents agreement. If the father agreed to give some money or agricultural land for the children, the mother will take them with her, otherwise the children will live with their father. Since both husband and wife have equal right on their house one of them will leave the house on agreement. (Yetmen)

 

However, this research did reveal that there were changes in place with the fall of the imperial and Derg era. New laws were imposed so that women could refuse the proposal to marry their dead husband’s brother, and that they could retain their rights to property. There was also a greater prevalence of women getting HIV/AIDS testing before marriage and greater tendency of couples marrying for love as illustrated below

 

There is still dominancy of men. There is no much change since 1991; women are still confined in the house. However the way of marriage changed a lot, the ceremony is now completed with a small expenditure and early marriage decreased highly, less divorce, widows are getting a good judgement from the court.  (Yetmen: WIDE2 2003)

 

In Korodegaga there was also evidence that couples were eloping to get married,  calling it ‘voluntary abduction’ to avoid arranged marriages.

 

4.4: Violence on women’s bodies

 

Unequal power relations are also expressed through physical violence towards women’s bodies in the form of female circumcision (known as one type of female genital mutilation (FGM)) and widespread domestic abuse. What makes this more alarming is the fact that woman accept this violence without question, thus highlighting the extent to which gender inequalities are firmly entrenched within society. Kabeer (2003) notes that power is most effective when it restricts choice and hence agency without overtly appearing to do so.  Indeed, a recent report by the World Bank (2005)[18] drew on data from a nationally representative household survey in 199