Betwixt and Between: Adolescent Transitions and Social Policy Lacunae in Ethiopia

Many adolescents in Ethiopia face difficulties in successfully transitioning to healthy adulthood. This can stem from challenges experienced by two unique phenomenon: ‘early adulthood’ where adolescents are forced into adult responsibilities too early; and ‘waithood’ where adolescents are unable to obtain the foundations they need to successfully move into adulthood. This paper uses the lens of these two opposing challenges to explore adolescents’ ability to successfully transition into adulthood and examine how far social policies support these transitions. It uses qualitative and quantitative data from the Gender and Adolescence: Global Evidence (GAGE) study on adolescents and young adults, their caregivers and key informants, from diverse rural, urban and pastoralist settings in Ethiopia. Our findings show that adolescent transitions are impacted by a range of diverse barriers with high levels of unsupported adolescents facing early adulthood or waithood. As such, we find that adolescents are being failed by social policies at each end of the spectrum. Social policies could provide a foundation to support these adolescents, yet we find that many are insufficiently age-disaggregated to take into account their diverse needs. In order to appropriately support these individuals in transitioning to healthy adulthood, social policies need to be targeted at adolescents throughout each stage of their transition. Furthermore, such comprehensive age- and gender-responsive social protection should be complemented with wider programming, such as awareness building, skills training, and adolescent-friendly health and mental health services.


Introduction
The period between childhood and adulthood presents a stage in life of semi-dependency where young people are no longer perceived as reliant children, yet do not receive all the benefits of fully independent adults. In Ethiopia, this transition is complex and varies substantially between different groups. While some adolescents are forced into accepting their adult responsibilities too early -often in the form of early marriage, parenthood or work -others enter into a period of 'waithood' in which young people are unable to transition into fully self-sufficient adults due to a dearth of livelihood opportunities and a shortage of assets to support an independent life (Honwana, 2014). Yet, as we will show, there is a lack of social policies targeted at individuals in this stage of adolescence, particularly that take into account early adulthood and waithood. This is problematic given the importance of adolescence -a period of substantial psychosocial and biological change -in the life cycle, as well as the substantial proportion of the population that youth comprise in Ethiopia, a country which typifies the growing concern with the so-called 'youth bulge' 1 in that around 25% of the population are between 10 -19 years old (PMA, 2020(PMA, , 2016. This article examines the ways in which adolescents navigate their pathways from childhood into adulthood within their social, cultural and structural contexts. It draws on qualitative and quantitative research from the Gender and Adolescence: Global Evidence (GAGE) baseline interviews with 6,826 adolescent boys and girls aged 10 -19 years, along with their caregivers and service providers, in communities from diverse rural and urban regions in Ethiopia (pastoralist Afar, Amhara, Oromia and Dire Dawa). It starts by conceptualising adolescent transitions, it then provides an overview of the social policy landscape in Ethiopia. Next, it discusses the results in relation to two unique challenges that adolescents experience when transitioning to adulthood: 'early adulthood' in which adolescents are forced into adult responsibilities early, and 'waithood' in which adolescents do not acquire the skills they need to become fully independent adults. The core findings are presented in relation to a number of interacting markers of adulthood which uniquely position adolescents' ability to successfully passage into adulthood: 1) movement through formal education, 2) development of an independent identity and voice, 3) entry into a separate family life, typically via marriage, 4) entry into parenthood, and 5) entry into paid work. We describe the risks adolescents face when they enter these stages either too early or too late. We then analyse the extent to which social policies take into account the changing and diverse experiences of adolescents at different stages of their transitions from childhood to adulthood. We conclude with the steps needed to create a comprehensive social policy environment in Ethiopia that will facilitate adolescent transitions along positive trajectories.
Our research was guided by two key research questions: 1) What are the key factors that hinder healthy and empowered adolescent transitions into adulthood? 2) How do Ethiopian social policies support adolescents through these diverse transitions and what are the key gaps?

Conceptualising Adolescent Transitions
Adolescent transitions into adulthood do not occur at one single point, but instead represent a series of events and subtle processes that are influenced by social, political and economic factors that shape adolescent trajectories (EGRIS, 2001). Thus, the boundary between childhood and adulthood is hard to define and adolescent transitions are becoming increasingly complex due to the vastly different routes adolescents can take to adulthood (EGRIS, 2001;Valentine, 2003). Beck (1992) suggests that this changing and diverse set of trajectories for many adolescents results in young people being unable to predict the future that awaits them. Adolescence is an important juncture in individual development and plays a pivotal role in shaping the direction their future lives will take. It is also a period of time in which social and gender norms become more entrenched and societal inequalities are reflected in adolescents' own future trajectories (Hunt, 2006). For girls in low-income countries, these gender norms tend to shrink their opportunities, as they are often forced to go down the preferred route of marriage and motherhood, yet boys typically see their opportunities widen (Harper, 2018). Adolescence presents a critical juncture in which social policies designed to combat these harmful social and gender norms could change the sorts of lives that girls and boys ultimately lead. These life-course events have an impact on adolescents' vulnerability to poverty (Moore, 2005). Thus, adolescence is increasingly seen as the 'age of opportunity' in which interventions can have drastic impacts on the potential to break out of intergenerational poverty cycles (UNICEF, 2011).
In a global trend, the delaying of common markers of adulthood -such as marriage, completion of education, and parenthood -has resulted in a prolonging of the transition from childhood to adulthood and the adoption of adult roles generally occurring at a later stage, with some experts calling for an expansion of the definition of adolescence to 24 years to account for this (Sawyer et al., 2018). This encompasses a period of semi-dependency where adolescents are no longer seen as children but yet do not receive all the benefits of fully independent adults (Valentine, 2003). It is increasingly common for adolescents to enter a stage defined as 'waithood', described by Honwana (2014) as the 'period of suspension between childhood and adulthood' where 'young people are no longer children in need of care, but…are still unable to become independent adults'. This is increasingly becoming a reality for many young people, due to poor education and economic opportunities which result in them being unable to attain the assets needed to gain social recognition as adults (Ibid.). Thus, adolescents enter into a period of stagnation where adulthood remains out of reach, which has implications on health and mental health (Bernays et al., 2020), economic empowerment (Honwana, 2014), and marriage practices (Singerman, 2007). Yet, this period is not defined as a passive state; those experiencing waithood often use their situation to actively interact with society in new ways (Honwana, 2012;Honwana & de Boeck, 2005).
By contrast, some adolescents can face truncated adolescent transitions, with their childhoods cut short due to an unsupportive environment that requires them to face adult responsibilities prematurely, such as entering into early marriage, parenthood or child labour (Honwana, 2014). These are influenced by a number of external factors such as poverty and social norms that value marriage and child labour over education which can make it more likely for some adolescents to experience early adulthood (Tafere & Tiumelissan, 2020). In addition, experiences of early adulthood and waithood may be discontinuous, whereby young people in early adulthood may later face periods of waithood (Honwana, 2014). While navigating these transitions, it is important to ensure young people are both not robbed of their childhood too early but equally are supported to acquire the skills they need to enter into adulthood at the appropriate time, and especially to transition effectively from formal education to paid work.

Social Policy Landscape
Ethiopia has an extremely high proportion of 15 -24 year olds in their population, as high as 21.7% in 2017 (Population Reference Bureau, 2017). Harnessing the potential of this 'demographic dividend' by supporting youth to actively participate in the workforce could reap substantial benefits (Pankhurst & Dom, 2019). Yet, unemployment is a key concern for many adolescents in Ethiopia (Broussard et al., 2014), forcing many into a period of waithood in which they are unable to enter the workforce and start their adult lives. Currently in Ethiopia youth transitions into paid work are slow and research evidence underscores that job opportunities do not match young people's skills or expectations (Pankhurst & Tafere, 2020). In rural areas, a change in livelihood opportunities, with a shift in focus away from agricultural activities -largely due to land shortages, climate change and population growth -coupled with insignificant job creation has contributed to low employment rates (Pankhurst & Dom, 2019). Such labour market changes have been a key driver of high rates of youth migration both domestically and abroad . These shifts have contributed to an altered experience of transitions for many adolescents in recent years.
While the reality described above is common for male adolescents in Ethiopia, adolescent girls often experience the opposite phenomenon in which they are forced into 'early adulthood' due to social and gender norms which place value on early marriage and childbearing. The most recent Demographic and Health Survey (DHS) showed that 40% of women aged 20 -24 were married before the age of 18, and 14% were married before the age of 15 (CSA & ICF, 2016). That being said, boys in Ethiopia can also see their childhood ending prematurely. This is often due to the pressures they face providing economically for the household. Ethiopia has extremely high rates of child labour, particularly for boys at 69% (CSA & ICF, 2016), which can result in boys having to face adult responsibilities early. Furthermore, child labour is linked to early school leaving and poor psychosocial outcomes (Sebany et al., 2019;Tafere & Pankhurst, 2015), largely impacting their adolescent transitions. Although significantly smaller than adolescent girls, a cohort of Ethiopian boys also enter into marriage as children -9% of men in Ethiopia have married before the age of 18 (CSA & ICF, 2016).
In Ethiopia, adolescent experiences of transitions are fast changing due to rapid urbanisation and significant socio-economic and cultural shifts (Young Lives & UNICEF, 2020). Commitments by the Ethiopian government to gain middle-income status by 2025 have driven a number of social policy initiatives -such as the Growth and Transformation Plan (GTP II) (2015/16-2019/20) -to lift many Ethiopians out of poverty. This has led to remarkable progress, with the poverty rate decreasing from 61 to 26% in 18 years (UNDP, 2015). However, these economic developments appear to be only benefiting specific parts of the population, with others, particularly in rural areas, benefitting very little which can further contribute to the experiences of 'waithood' (Abebe, 2020). Within the past two decades, the budget allocation for schools has doubled (World Bank, 2018). Prioritising education has had positive impacts on enrolment rates, with the number of primary school students increasing from 3 to 21 million in 20 years, resulting in an expansion of opportunities for many adolescents (MoE, 2017).
Ethiopia has introduced a number of national policies aimed at expanding the opportunities of young people. In 2005, the National Youth Policy was introduced, bringing attention to a number of key objectives for youth, including creating an empowered young generation, increasing youth participation in formal and informal employment, and improving unbalanced educational participation in rural and urban settings. More recently within education, the Ethiopian Education Development Roadmap highlights key areas that require fundamental change within pre-primary, primary, secondary and higher education for 2018 -2030, in particular, a lack of quality learning outcomes, inadequate educational facilities and poor quality teaching. Ethiopia has also made impressive efforts to combat child marriage through the introduction of a number of policies 2 and the creation of the National Alliance to End Child Marriage and FGM/C. Additionally, it is one of 12 countries supported by UNFPA and UNICEF's Global Programme to Accelerate Action to End Child Marriage (UNFPA-UNICEF, 2019). Correspondingly, the average age at first marriage has climbed, although there is still significant regional variation (Central Statistical Agency (CSA) & ICF, 2016). Yet, it is estimated that progress needs to be 10 times faster to eliminate child marriage by 2025 (MoWCY and the National Alliance to End FGM & ECM, 2019).
In order to successfully support adolescent transitions, social policies should take into account the varying and diverse trajectories discussed above, but in many contexts they fail to reflect the increasingly de-standardised routes towards adulthood (EGRIS, 2001). This paper aims to understand these transitions and the extent to which social polices both support adolescents who are forced into early adulthood and conversely those who experience prolonged adolescence in the form of waithood.

GAGE Study Design, Ethics and Conceptual Framing
This paper draws on mixed-methods research conducted in 2018 for the GAGE multi-country longitudinal study, which follows 20,000 adolescents from the ages of 10 to 19 years over a 9 year period in 6 different countries. This paper uses data from adolescent boys and girls in Ethiopia in order to study the specific factors that shape and truncate adolescent transitions into adulthood. The quantitative survey includes data from over 6,826 adolescents and the qualitative data includes a sub-sample of 240 adolescents. A subset of the quantitative and qualitative dataset was purposively sampled to include adolescents with disabilities and early married adolescents.
The study is informed by GAGE's conceptual framework, which takes a holistic approach and highlights the interconnectedness of 'the 3 Cs' -adolescent capabilities; programmatic change strategies; and contexts -in order to effectively understand 'what works' in empowering adolescent boys and girls (See Fig. 1). This framework builds on the capability approach coined by Amartya Sen (1984 2004) and later further developed to include gendered realities at the household and societal levels (Kabeer, 2003;Nussbaum, 2011). It highlights the need to invest in various social, economic, human, political and emotional assets for individuals to able to achieve valued ways of 'doing and being'. In our framework, these capabilities are defined as: education and learning; bodily integrity and freedom from violence; health, sexual and reproductive health and nutrition; voice and agency; psychosocial wellbeing; and economic empowerment. It is recognised that these assets are situated within an individual's context at the household, family, community, state and global levels. These contexts will further influence change strategies that can be implemented to support adolescents. For a transformative approach to occur, they must be integrated at all levels, working with policies and programming to support adolescents and their families and communities.

Data Collection
We explore Ethiopian adolescent transitions in diverse urban, rural and pastoralist contexts in the regional states of Amhara, Oromia and Afar, and in the Dire Dawa city administration. The quantitative sample includes two cohorts: younger 10 -12 year old (n = 5,495) adolescents spread across the research sites, and older 15 -17 year old adolescents in urban areas only (n = 1,331). (see Table 1 for an overview of the quantitative sample). The qualitative sample includes younger adolescents (10 -12 years) and older adolescents (15-19 years) in all research sites in order to explore experiences of early adulthood, complemented with additional young adult (aged 18-30) focus group discussions to feed into the discussion on waithood. The qualitative data includes 240 adolescent individual interviews, which were complemented by 48 focus group discussions (FGDs) with young adults, 15 social norm focus groups with parents, 33 community mapping focus groups with adolescents, 195 parent interviews and 174 key informants (see Table 2 for an overview of the qualitative sample).
Data was collected in 2018 through in-depth interviews and focus group discussions using a range of participatory research tools, including community mapping and 'my favourite thing' exercises, designed to start conversations and explore more sensitive topics (for more information on the qualitative tools used see Jones et al., 2019;Jones et al., 2019). Data was collected by local researchers with regional language skills. To ensure participants were comfortable discussing sensitive topics, female researchers conducted interviews with female participants and male researchers conducted interviews with male participants. Focus groups were conducted in single sex groups in order to facilitate conversations between participants as girls and women may be less likely to actively contribute in mixed sex groups due to conservative gender norms.
Ethical clearance was obtained for research activities from the Overseas Development Institute (ODI) Research Ethics Committee (02,438), the Addis Ababa University College of Public Health Institute Review Board (113/17/Ext), and the Ethiopian Development Research Institute (EDRI/DP/00,689/10). The studies were also approved by regional governments in each regional state.
Adolescents over the age of 17 gave their consent, while assent was obtained and consent was provided by parents or caregivers of younger participants, except in child-headed households. The option for oral assent or consent was provided for illiterate participants or caregivers. Any safeguarding issues that came up in the interview were flagged by the researcher, and participants were referred to the appropriate services. All participants were provided with a phone number to raise any concerns.

Data Analysis
For analysis of the qualitative data, transcripts were translated into English by native speakers of the local languages and data was coded thematically by a team of experienced coders using the qualitative software MAXQDA. The codebook was developed iteratively and informed by our conceptual framework (GAGE consortium, 2019), as well as by researchers in the field to add local specificity. Analysis was undertaken by experienced researchers using a capability lens to highlight particularly prominent truncated adolescent transitions in each of the six capabilities, and results were explored in relation to gender, age and location specific challenges. For analysis of the quantitative data, key outcomes of interest within each of the six capability areas were constructed using the individual-level survey interview information. Annex 1 of GAGE's conceptual framework (GAGE consortium, 2019) maps out all the indicators -which are operationalised in the quantitative survey (Baird et al., 2019). Outcomes were then summarised by gender, age group and residential location.

Results
We find that adolescents experience a number of challenges to successfully transitioning into adulthood. The challenges range from adolescents being forced into early adulthood through child marriage, child labour and parenthood. Yet, older adolescents face unique challenges due to a lack of education and employment options, preventing them from transitioning into independent adulthood. In both these instances, these experiences prevent adolescents from reaching their full capabilities. In each of the areas explored, social policies could present a strong foundation to support adolescents in breaking away from these pathways. Yet many gaps persist, and polices are not sufficiently age-disaggregated, resulting in adolescents receiving inadequate support through their transitions.

Child Work
Early adulthood can prevent adolescents from access to decent and age-appropriate employment. During their early years, adolescents from poorer households often find themselves having to take on financial responsibilities prematurely in order to help provide for their household. Although the legal age of employment in Ethiopia is 14, many employers in rural and urban areas ignore these legal restrictions. Our quantitative survey finds that 5% of the younger cohort and 21% of the older cohort are engaged in paid work. The qualitative data show many underage adolescents working full time in informal, manual labour. Adolescents are engaged in different types of work, including domestic and farming family labour, paid work in factories and cash crops farms and this can have implications on the health and wellbeing of adolescents. One 12 year old girl from South Gondar, Amhara describes finding work when she was 9 years old in order to help provide for her family: 'There is a guy who travels by motorbike….I told him that I want to work, then he found me the job…. We just bring small rocks and then put it somewhere and he pays us….
[I choose to work]… to change our life and get food to eat…. They don't tell you that you can't work because you are just a child'.
In Batu town, large factories such as Flower Company A 3 make up the majority of the employment options. Although the minimum age for working in the factory is 18 years, 4 we found adolescents as young as 13 are working there for up to 12 h a day. As a participant of a focus group discussion with migrant adolescent girls in Batu, Oromia, explains: 'I left to Ziway in search of work. I was just a kid and even the people there didn't believe I was fit to work. I lied that I was 15 when I first applied for the job. I wrote 15 on my ID with a pen'.
The forms of child labour that adolescents are often engaged in typically consist of long hours with risks of exploitation from their employer. In East Hararghe, farming of the seasonal crop Khat is a common economic activity for young adolescents. As an 11 year old girl from East Hararghe, Oromia explains, the working conditions are often extremely difficult: 'They don't even allow you to eat your food. They rush you to finish your food and return to the work. Due to this, we are sometimes forced to eat when we are about to leave for home…the work is difficult because we cannot break until it is time for lunch… there are also some employers who deny your payment'. The quantitative data finds that adolescent boys were more likely to be engaged in paid work than adolescent girls; girls were 22% less likely to be in paid work in the younger cohort and 50% less likely in the older cohort (see Tables 3 and 4). 5 While boys are primarily pulled into formal forms of employment, adolescent girls often see their childhood truncated through unpaid work in the household. As a participant in a focus group discussion with fathers from East Hararghe, Oromia, explains: 'A girl has to be absent from the school to take care of the small child until her mother returns from fetching the water. She cooks, and feeds the child'. The unequal distribution of household chores can result in poorer school attendance among girls and a lack of time for homework, causing them to fall behind in school and limiting their educational trajectory.
In Afar, where pastoralism is the primary livelihood option, young adolescents are relied upon by their families to undertake herding activities. The responsibility of the household is placed on these adolescents at a very young age, leaving them no time to engage in activities that are usually associated with their age group, such as attending school or playing with their friends. As an 11 year old boy from Afar who spends his days herding, alone or with his brother describes: 'We have temporarily migrated to a locality, which is far from our locality and we are residing there…There are no children herding livestock with me; rather I spend the whole day alone in a forest. I prefer watering the young camels in the morning and spending my time at home after that, but my father wants me to practice herding camels and to be able to take full responsibility for taking care of the camels in the future'.

Education
Education and learning has a significant influence over adolescent transitions and is a vital component for adolescents to reach their capabilities. Yet, as adolescents in Ethiopia transition through their teenage years, it becomes increasingly difficult to stay in school, particularly for girls. This results in many adolescents' educational trajectories being cut short, as they often move straight into paid work or marriage and parenthood. Our quantitative survey finds slight gender differences in the younger cohort, with boys (86%) more likely to be enrolled in school than girls (82%). Although this is beginning to change, our qualitative findings highlight that gendered social norms can result in many girls leaving school as they reach mid-adolescence. As one key informant from East Hararghe explains: 'School dropout starts from grade 8 exacerbated by low economic status of the parents to pass their children to the next grades…Many girls dropped out of school compared to boys because of marriage and economy'.
Transitioning through education is also not straightforward for adolescent boys, and as described above, many are pulled out of school to engage in paid work due expectations around providing for the household. As a 17 year old boy from Dire Dawa explains: 'I was in grade 8 [when I quit school]. My mother began to lose in her business so that I had to start work in order to help her. Her expenses became higher when I go to school, I take some money for a Bajaj and there is breakfast and other meals for me and my sister. The only option that I had was starting a job'.
Although the rates of children enrolling in school are increasing each year, age of enrolment is a key issue. Our quantitative survey finds that the average age of enrolment in grade 1 was 7.7 years. However, this differs regionally, with Afar having the highest average age at 8.2 years. Our qualitative work highlights that this is often due to responsibilities related to herding livestock or the lack of school infrastructure in the region. A 12 year old girl from Afar explains: 'It varies. [You might join grade 1] at age 11 or 12 or 13….I didn't start grade 1 before age 11 [because] they didn't send me to school…..I was keeping baby goats'.(see Tables 5 and 6) Learning outcomes in Ethiopia are still extremely poor, particularly in rural areas, our qualitative data suggests that this is largely driven by poorly resourced classrooms, unmotivated teachers and limited parental support for education. Indeed, many teachers highlighted a significant lack of teacher training, leaving them unprepared for the classroom. As Table 4 Economic Empowerment, Young Cohort, Gender and Disability Table 3 Economic Empowerment Old Cohort, Urban Only a school principal from East Hararghe highlights: 'There is no training given on how the teachers should handle the students. There is also lack of materials…there are not enough classrooms and playgrounds. So it is below the standard'. In remote rural areas of Afar, where access to classrooms and teachers is largely non-existent, these structural barriers can have striking effects on adolescents' ability to continue their schooling.
A prominent barrier to older adolescents' educational transitions is the lack of accessible secondary schools, especially in rural areas. Adolescents in remote communities often have to travel long distances to attend school post grade 8. This mostly impacts girls, who are at risk of sexual assault on long journeys to and from school. As a group of fathers from South Gondar explains: 'Parents are not interested to send their daughters to high school. This is because of the fact that girls in our community faced sexual assault when they travelled back and forth to high school, which in turn forced them to drop out of school'. Many adolescents stay in rented accommodation near the school to mitigate these risks, but they need the financial resources or family contacts to enable this, or they end up dropping out (Devonald et al., 2020).

Early Marriage and Childbearing
Whereas boys often find their youth curtailed by entry into paid work, we find that early marriage is a major cause of girls having to face adult responsibilities prematurely. Those who marry as children see their opportunities diminish; they are much less likely to be in school, have lower expected earnings, and are more likely to be young mothers (Wodon et al., 2017). However, our qualitative findings highlight that for some adolescents, particularly those in Amhara, child marriage is becoming increasingly uncommon, and adolescents are able to decide when and who to marry. As a father of an adolescent in South Gondar describes: 'Yes, they are now refusing to get married with the force of their par-ents… because they go to school now, they realise their rights so they are also refusing to get into an arranged marriage'. Historically, Amhara had the youngest median age of marriage, so remarkable progress has been made in raising this age and tackling the most harmful forms of child marriage.
However, in the two other study regions (Afar and East Hararghe), the situation is markedly different, and child marriage is still common. In Afar, child marriage rates are still extremely high and often enforced by parents through the absuma system (marriages are arranged to maternal cousins and adolescents generally have no say in who and when they marry). There is strong support for this system in the community from both parents and some adolescents (boys are particularly supportive of the practice) due to the reinforcement of kinship ties. As such, going against the absuma system is met with severe punishment. As a 13 year old girl from Afar explains: 'In our culture, parents arrange marriage and tell to the girl suddenly. I was angry when I heard as I am going to get married, but I calmed down since it is the culture'. Additionally, whereas boys can wait until early adulthood to marry through the absuma system, girls are often made to marry in mid-adolescence, leave school, and take on responsibility for the household, rapidly cutting short their adolescence.
In East Hararghe, progress on eliminating child marriage is similarly thin, with the median age of marriage remaining unchanged over the last two decades (CSA & ICF, 2016). Child marriage has always been strongly engrained in the community, and many parents in this area pressure their daughters to marry young. However, there has been a surprising trend in adolescent-driven marriage in this area. This is often due to stigma that results in older, unmarried girls being described negatively as 'haftu.' However, the picture is a lot more complex than this, and other drivers, such as a lack of education or work opportunities (that leave many adolescent girls with a dearth of alternative options), or the influence of peers marrying early, play an increasingly prominent role, as a man from a community timeline discussion from East Hararghe explains: 'As a result of drought, families, due to economic problems, are unable to buy a pen let alone buy other necessities… Girls prefer marriage than to simply sit idle'.
Furthermore, adolescent girls face significant pressure to get pregnant shortly after marriage in order to demonstrate their fertility. As a result, many find themselves facing early motherhood, ensuring an abrupt end to their adolescence as they take on child caring responsibilities. The focus on childbearing for married adolescents results in a lack of knowledge about, and stigmatisation of, contraception, as a 14 year old girl from East Hararghe explains: 'I am not using [family planning] now -before I have one child. If you stay without a child for a longer time, they will tell you, you are barren'. This is reflected in our quantitative survey, with only 25% of the younger cohort able to identify a form of modern contraception. This differs significantly by location with the lowest levels of knowledge seen in Afar (7%), followed by East Hararghe (14%), and South Gondar (40%).(see Table. 7) The consequences for adolescents who face premarital early motherhood are far greater. Adolescents who find themselves pregnant out of wedlock face severe discrimination from both their family and the community and often have to drop out of school due to the stigma and unsupportive environments they face. As a group of 15 -16 year old adolescent girls from South Gondar explains: 'There is this girl who was grade 6 last year…she got pregnant while she was grade 6…They'd call them names that are demoralizing, like 'You whore!', 'You are useless'. They'd blame them like that. Her family became too upset and forced her to drop out of school, denying that she is their daughter'.
For unmarried adolescents, using contraception has additional challenges and is often highly stigmatised. However, in Amhara, contraception is beginning to be discussed more freely and has higher rates of use. This may be due to fears of the physical, social and economic implications of early pregnancy, as well as fears of becoming pregnant through sexual violence, as described by a mid-adolescent girl from South Gondar: 'Girls that don't have a lover get injected, fearing rape and getting pregnant'. In Afar, the availability of contraception is also increasing slightly, due to changing sexual norms in the region. Adolescents taking part in the cultural dance 'sadah' appear to be increasingly involved in sexual activity, and thus unmarried adolescents are beginning to secretly use modern contraception. As a girl from a 15 -17 year old community mapping exercise explains: 'The unmarried girls are using the medicine until they made their marriage legally. So we are using…the contraceptive methods…for three or six months'.

Economic Empowerment
At the other end of the spectrum, older adolescents in Ethiopia can find themselves unable to acquire the assets they need to become fully independent adults. Limited economic options have resulted in adolescents entering a period of extended waithood. In our qualitative work, respondents consistently reported high youth unemployment rates, as a key informant from Dire Dawa highlights: 'The major challenges that adolescents encounter Table 7 Health, Nutrition, Sexual and Reproductive Health, Young Cohort, Location in recent years is joblessness. It is both dropouts and graduate adolescents that experience unemployment. One of my relatives is a university graduate in science, but he wasn't able to find a job'. It can be even harder for certain groups to find employment or training opportunities. In our quantitative survey, we found that adolescents with disabilities in the older cohort spent 31% less time in school, training, and employment compared to their peers (25% vs 17%).
Expanding effective job creation programmes targeting youth and those with limited education is a vital component in tackling waithood. However, job creation programmes targeting adolescents are rare, and often exclude the most vulnerable groups (for example, those who have not completed grade 10). In urban areas, job creation and credit programmes are more common. However, there are reports of corruption within these programmes and difficult registration requirements for younger job seekers. As a group of 16 -17 year old boys from Debre Tabor, Amhara explains: 'One of the barriers was our age. Most of us are below 20 years but to organise in group the government require you to be above 20 years….. I was going to Kebele to organize and work in cobble stones. They requested to renew my Kebele ID card. However, they asked 80 birr to pay to renew my ID so since I had no money at the moment the registration time passed''.
Loans could provide adolescents with the resources they need to move out of waithood, yet due to the collateral needed to acquire loans, they are either reluctant to take on risk, or do not have the necessary assets to secure a loan. As a 17 year old boy from Batu explains: 'I heard there is saving and credit institution in the town and some adolescents use credit service. But, the service is not easily obtained, as it requires delivering title deed as collateral to get credit service. It needs only one member of the group to deliver title deed, but in case the money obtained through credit is wasted, it is up to the individual who provided the title deed to pay the money. Thus, people fear to deliver a title deed to get credit, thus majority of the youths are not accessing the credit service'. Additionally, we find gender differences in accessing credit services; whereas some boys are supported in get their driver licence for work as Bajaj drivers (three-wheel taxis), opportunities for girls are more limited. Our quantitative data further highlight gender differences, with girls in our older urban cohort 47% less likely to have money they control compared to their male peers.
The lack of employment options has driven many older adolescents to migrate in search of better job opportunities. As a key informant from East Hararghe explains: 'They are migrating to win life. In our locality there is no job opportunity. Their only option is migration'. In Amhara, boys often migrate to the lowland areas of Humera and Metema for seasonal work on farms in the hope of earning enough to send money back to their families. However, precarious working conditions, low wages and unreliable work means they can rarely achieve this. Furthermore, the limited job opportunities for adolescent girls and young women has seen a dramatic spike in them migrating internationally to work in domestic service -often in neighbouring Arab countries and under highly risky and exploitative labour conditions (Jones, Baird, et al., 2019). However, some Arab countries have started deporting Ethiopian domestic workers, which has resulted in adolescent girls returning to Ethiopia with even fewer economic opportunities.

Education
A significant number of adolescents fail grade 10, which can result in them leaving education without the tools they need to successfully transition into adulthood. The lack of quality teaching in many schools can result in adolescents falling behind and being unprepared for the national exams required to move onto higher levels of education. As such, many are unable to continue into tertiary education, such as university or technical college, as a group of 18 year old girls from Debre Tabor describe: 'We have problems in schools. For example, the textbooks are not sufficient… The number of students attending here are many. The teachers are not well qualified and because of such shortages the probability a student passes grade 10 is very small'. Failing these exams reduces adolescents' future opportunities and drives early marriage and migration for work, or results in them remaining at home without work. As an 18 year old girl from Batu states: 'My older brother was a clever student but failed from the 10th grade. He is now at home doing nothing'.
That being said, university qualifications do not guarantee job prospects, and there are high unemployment rates for graduates. A group of 17 -19 year old students from Batu explains the situation: 'You see university graduates sitting at home unemployed. It is so disappointing and makes you feel hopeless. It makes you worry what our fate is going to be after we graduate from university'.
The lack of job prospects for graduates can have knock-on effects on educational aspirations in the community. Many become disengaged in their future or no longer see the benefits of education, as a group of 15 -16 year old boys from Debre Tabor explains: 'Due to large number of unemployed graduated youth there are families who are demotivated to send their children especially girls into school. Most of the youths are also migrating from the rural to the urban areas when the situations there are not conducive for them'.

Psychosocial and Other Health Risks
High unemployment rates can have an impact on adolescents' psychosocial well-being, resulting in a feeling of helplessness which can encourage some adolescents to engage in risky behaviour. Substance abuse, largely driven by a lack of employment options, is increasingly becoming a major problem for adolescent boys in Ethiopia. A group of older adolescent girls from Dire Dawa explains: 'When you are hopeless with life, you slowly settle in your living situation. That means when you spend time with people like you, you will become idle and engage in khat addiction which is very common in the area. There are many people who involve in this because of friends pressure'. High rates of substance abuse have been linked to significant levels of peer violence, and generally cause adolescents to further disengage from their future, severely impairing successful transitions to heathy adulthood. As a father of an adolescent in South Gondar describes: 'Male adolescents disturb when they get drunk by fighting each other….Such fighting took place by using stick and even using a knife. It is a recent phenomenon….It is mostly happen while they met in a drinking house……These are individuals whom they couldn't get a pass mark in their exam. They often fought each other and disturb others, as they are jobless'.
This period of waithood may also result in adolescents taking up precarious and often dangerous work due to a lack of other options, which can have resultant impacts on their health. This can be seen in Batu, East Hararghe, where adolescents engage in work in Flower Company A -a factory known in the community for exploitative and dangerous working conditions -often as a last option. An older boy from a community mapping exercise explains: 'Out of the total job opportunities in the town, 75% is in the Flower Com-pany…..it has a lot of problems, particularly the chemicals used in the flowers is affecting their health. They are doing it because of lack of access to other job opportunities. Those who have at least some start-up money don't engage in this job rather they work on their own petty trading'.
In a similar vein, adolescents who take up precarious day labour in the highlands region are also exposed to serious health risks. This region of Ethiopia experiences high rates of malaria and the working culture can foster unhealthy lifestyle choices. Risky sexual practices-often stemming from high levels of substance abuse -which result in low condom usage-has resulted in high rates of HIV, as a male community group in South Gondar explains: 'The boys are involved in the cash crop farming in Metema. However, the majority of the boys get sick and the parents are forced to expend money to take back their children'. Although these health risks are well known in the community, adolescents still migrate from all over, as a group of young men who migrated to Metema in South Gondar explains: 'Many of them migrate because there is no option here. Even if they know death is ahead of them, they go there thinking it is better to die eating than sitting here doing nothing'. Adolescent boys also face multiple threats to their bodily integrity while living in lowland areas, as a result of general lawlessness and low policing. Rates of violence are extremely high, especially between groups of boys from different areas, who fight over work, wages and girls.
Adolescent girls who migrate to urban areas or abroad to the Middle East also lack protection mechanisms. Domestic workers can be exposed to predatory employers, risking sexual violence, and in urban areas of Ethiopia, coercion into commercial sex work, as a 17 year old adolescent girl participant in a focus group in Dire Dawa explains: 'They say "you will work as a maid", but then they are commercial sex workers when they come here. They can't change work because they become addicted'.

Voice and Agency
Aligning with Honwana's (2014) findings on the impact of waithood on adolescent involvement in political social movements, we find that some adolescents actively set out to change their situation by demanding better job opportunities. A key informant from Bahir Dar, Amhara describes the context as follows: 'At a national level, the youth, especially jobless high school graduates, are demanding the government to create jobs and that is why the government started to give special focus to youth issues'.
In some cases, this has resulted in adolescent voices being taken more seriously at the community and national level, with government structures becoming more youth inclusive. A group of 17 -19 year old girls in Batu describe the changing situation: 'The past two years, there have been unrest. Questions were raised by the community. I think the government is now opening its doors to the youth. They are calling 17 and 18 year olds to meetings. They are asking them for suggestions. But these meetings are usually for university students who come back home during summer'.
Our quantitative survey, however, shows that there is a large gender difference in older adolescents' participation in the community. Boys are 49% more likely to have taken joint action for a serious problem affecting the community than girls, and are 55% more likely to have talked with people about a serious problem affecting the community.(see Table 8) In Oromia, men and boys have formed young activist groups called 'Qeerroos' at the community level in response to recent political tensions in Ethiopia. They are engaged in helping youth stay in school and provide aid to the large populations of internally displaced people in the region, a group of 16 -18 year old male Qeerroo members from East Hararghe describes: 'At this time 'Qeerroo' has been established in all the neighbourhoods in this kebele… We are working in creating unity through working with different communities.
We are supporting poor people through constructing houses for them… collecting contribution for displaced people [and] are in charge of returning students who dropped out of school'.

Social Policy Gaps
Our findings indicate that in order to support a successful transition through adolescence into adulthood, social policies need to be targeted at adolescents at each stage of their transition. However, the 2004 National Youth Policy of Ethiopia (MYSC, 2004) lacks age disaggregation, broadly defining youth as aged 15 -29. Consequently, it does not take into account the differing needs of young people at each end of this age bracket. Furthermore, Ethiopia's Growth and Transformation Plan II (GTP II) (MoFED, 2015) aims to put in place the structural changes needed to accelerate growth and allow Ethiopia to become a middle-income country by 2025. Harnessing the economic potential of the so called 'youth bulge' is essential to making this a reality. Although the policy sets out to 'increase the beneficiary of youth from 10 to 30 percent by engaging 4.32 million youth in agriculture and 3.64 million youth in non-agricultural activities' (pp.210), it is not age-disaggregated and does not sufficiently take into account the varying needs of adolescents throughout the different stages of their transitions. In particular, it does not highlight the unique challenges of those who may no longer be defined as adolescents, but are experiencing waithood and require the appropriate support to allow them to enter adulthood.
Additionally, although legislation is in place to prevent younger adolescents from being forced into child labour, age restrictions for paid work are often not enforced on the ground, as our findings from the floriculture industry underscore. Additionally, domestic work is not governed by Ethiopian labour legislation, thus policies need to ensure the protection of domestic workers and international and domestic migrants. Within education, there has been a focus on improving access for children at primary level. The Ethiopian government has responded to low enrolment rates by rapidly expanding primary school education. Yet, although budget allocations have doubled, there are significant quality deficits (Woldehanna et al., 2018). Furthermore, the same financial investments have not been made within secondary schools. Social policies are beginning to recognise this gap and some are now seeking to 'increase expansion of secondary schools' (MoFED, 2015 pp.188). However, as seen with primary schools, increasing the number of schools with little concern for quality can result in illequipped schools and non-conducive learning environments. Attention therefore needs to focus on quality of instruction. As laid out in the recent 'Ethiopian Education Development Roadmap' (2018 -30), the introduction of teacher management systems aimed at monitoring the performance of teachers could be one way of achieving better quality of learning in Ethiopian schools (Teferra et al., 2018).
Education policy concerning adolescents in Ethiopia is largely focused on progressing to higher secondary and tertiary levels. However, there is less focus on supporting adolescents in gaining the vital skills they need for employment. This results in less attention on those at the other end of the spectrum, who have made their way through education but are still unable to enter employment.
Ethiopia's flagship public works programme-the Productive Safety Net Programme (PSNP)-is contributing to a reduction in poverty and some evaluation evidence suggests that it is improving adolescent girls' nutrition and education outcomes (Favara et al., 2019;Gilligan et al., 2009). However, there are significant concerns that in some households, girls' workloads have increased as they are forced to take on domestic work while their mothers participate in the programme, and some boys engage in PSNP work activities instead of their parents and may prevent some from attending school (Jones et al., 2019).
Due to growing health concerns, social policies are also targeting rates of substance abuse, mainly through school-based awareness sessions which effectively exclude the most at risk adolescent boys -those out of school and migrant workers (Jones et al., 2019). Stronger efforts have been made to reduce early parenthood by improving sexual and reproductive health, particularly through the effective Health Extension Workers multipronged strategy aimed at increasing awareness and access to contraception. This has been particularly effective in Amhara. However, challenges surrounding social and religious barriers in Oromia and Afar need to be addressed, and services need to be greatly expanded in Afar -an area in which functional services are severely lacking (Ibid.).
The decline in child marriage in Amhara has been linked to policies that are starting to have positive effects, such as school-based girls clubs, and community engagement that highlights the health risks of child marriage (Presler et al., 2016). However, some policies do not target context-specific practices and norms and lack attention to the specific regional differences. Tailored implementation plans are therefore vital in effectively supporting adolescent transitions.

Conclusion
Our findings show that adolescent transitions are impacted by a range of overlapping barriers which can prevent young people's realisation of their full capabilities, including education and learning; bodily integrity and freedom from violence; health, sexual and reproductive health and nutrition; voice and agency; psychosocial wellbeing and economic empowerment. For early adulthood, a range of overlapping barriers can push adolescents into exploitative forms of child work, force them to drop out of school and/or marry at an early age. These outcomes truncate adolescent transitions into adulthood and present challenges for reaching their full capabilities. However, older adolescents in Ethiopia may also find themselves unable to obtain the assets they need to become fully independent adults. Due to a dearth of employment options, many adolescents-even those who have university degrees-face unemployment and a lack of training opportunities. This inability to reach independent adulthood can have implications for adolescents' psychosocial well-being and result in them undertaking risky behaviour such as substance abuse and risky sexual behaviour. However, aligning with previous research (Abebe, 2020) some adolescents' experiences of 'waithood' has led an increase in their voice and agency through involvement in political social movements.
Overall, social policies are failing adolescents at both ends of the spectrum; those who are forced into early adulthood, and those who lack the skills and resources to move into adulthood. Both are not being appropriately supported. The multiple and cross-sectoral challenges facing adolescents in Ethiopia need to be addressed through coordinating mechanisms that can ensure a cohesive social policy landscape. Furthermore, policies should be age-disaggregated in order to take into account the varying needs of adolescents at each stage of their transition from childhood to adulthood.
There is a need for comprehensive age and gender-responsive social protection, such as cash and/or asset transfers for education and training, to support adolescents at each stage of their transition. These services should be complemented with wider services such as adolescent-friendly health, mental health and violence prevention schemes. In particular, tailored substance abuse services are strongly needed for the unemployed and migrant workers to tackle this growing epidemic.
Stronger efforts in schools and universities are needed to equip adolescents with employment skills and connect students with employers. Improving access and awareness of Technical and Vocational Education and Training (TVET) as a tertiary education option, alongside efforts to link course offerings to market demands, would also be beneficial. Community and parental outreach sessions should highlight the benefits of formal as well as technical and vocational education over child labour and marriage.
Finally, the Ethiopian government, with support from development partners, is at the early stages of rolling out a new community-based social worker scheme involving graduates who act as caseworkers for vulnerable children. This has the potential to support adolescents at each stage of their transition, both by training up graduates and by supporting vulnerable younger adolescents at risk of entering early adulthood and realise their full capabilities.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.