Child Protection and Social Inequality : Understanding Child Prostitution in Malawi

This article draws on empirical research to develop understandings of child prostitution, previously theorised on the basis of children’s rights, feminist, and structure/agency debates, largely ignoring children’s own understandings of their involvement in prostitution. Conducted in Malawi, which is one of the economically poorest countries in the world, the study goes to the heart of questions of inequality and child protection. Within a participatory research framework, nineteen girls and young women used visual methods to generate images representing their experiences of prostitution. Individual and group discussions were used to illuminate the meanings and significance of their images. With the exception of the youngest, participants understood their initial involvement in prostitution as a means of survival in the face of poverty and/or parental death, or escape from violent relationships, experiences that were subsequently mirrored by exploitation and violence within prostitution. Using the lens of the capability approach, we capture the complexity of child prostitution, demonstrating the ambiguous agency of participants in the face of deeply embedded patriarchal cultural norms that constrained their choices and limited their freedom to pursue valued lives. We end by reflecting critically on the theoretical and methodological contributions of the study, making policy and practice recommendations and identifying opportunities for further research.


Introduction and Background
This article focuses on how complex intersecting inequalities (Walby et al. 2012), based on poverty gender and age inform questions of child protection for girls that are involved in prostitution in Malawi.It presents an empirical study designed to develop a nuanced, contextualised understanding from the perspectives of those with experience of involvement in prostitution as children.

Poverty and Politics
Malawi, a landlocked country in Southern Africa, is ranked 170 of 188 countries on the Human Development Index compiled from a range of health, education, income, and gender inequality indicators (UNDP 2015a).Key socio-economic and health challenges include high levels of poverty with nearly three-quarters of the population subsisting on less than $US1.25 per day, and two-thirds experiencing multi-dimensional poverty that reflects access to health services and education, as well as household income (UNDP 2015b).Malawi is a very 'young' country with almost half its population being aged under 15.A high adolescent fertility rate linked to early marriage 1 is associated with

Gender Inequalities
Gender inequalities in Malawi are rooted in patterns of social interaction that characterise the daily lives of women (Minton and Knottnerus 2008, p. 182).While economic factors and poverty play a central role in constraining women's choices and opportunities, long held and deeply embedded cultural practices continue to prescribe and proscribe roles and behaviours of women and girls, limiting their access to key resources and perpetuating gender-based inequalities in the home, in school, and in employment.This is despite substantial investment through bi-lateral aid, as well as national and international non-government organisation (NGO) programmes that are designed to address gender inequalities (Moser 2005) including gender awareness training (Tiessen 2004).A notorious example is the USA funded Girls' Attainment in Basic Literacy and Education (GABLE) programme that introduced free primary education and a fee-waiver for girls in secondary schools in the 1990s.But, overlooking structural specificities that led to opposition from some sections of society, and attitudes to external interference in policy making, initiatives that were designed to improve girls' access to schooling were not fully internalised within the Ministry of Education.Under these circumstances, parents needed little persuasion to maintain the status quo that prioritised boys' over 2 42% of ever married women experience emotional, physical and/or sexual spousal violence, 34% sustaining injuries (National Statistical Office and ICF 2017).girls' education (Kadzamira and Rose 2003).As Moser (2005) has argued, while a gender discourse and gender policies have clearly been established, gender mainstreaming has been subject to 'policy evaporation' and the 'invisibilisation' of successes, symptoms of the resistance mounted by powerful political interests to limit the effectiveness of programmes channelled through the Ministry of Gender (Moser 2005).As Cornwall and Rivas (2015, p. 396) have commented in a global context, 'gender equality' and 'women's empowerment' are terms that have been eviscerated of conceptual and political bite, compromising their use as the primary frame in demanding rights and justice.

Child Protection
Malawi ratified the Convention on the Rights of the Child (United Nations 1989) in 1991, and the Optional Protocol on the Sale of Children, Child Prostitution, and Child Pornography (United Nations General Assembly 2000) in 2009.These commitments provide a basis against which progress in providing for basic survival and development needs, in protecting children from abuse and exploitation, and in promoting the participation of children in decisions that affect their lives can be evaluated (Jones and Welch 2018).Malawi is also a signatory to the African Charter on the Rights and Welfare of the Child that was brought into force in 1999.Advocating respect for African cultural values (Olowu 2002), article 31 of the Charter refers to children's responsibilities to contribute to family, social, and national solidarity, to preserve and strengthen African cultural values, and to contribute to the moral well-being of society.
As the notion of children's rights and respect for African cultural values are held in tension, childhood in Malawi continues to be strongly informed by a patriarchal culture (Tiessen 2008) that locates power on the basis of gender and age.As outlined in Section 1.2, boys are prioritised over girls in accessing education and girls' access to education is limited by other factors, including experiences of gender-based violence that discourage attendance, even at primary school (Bisika et al. 2009).In the broader context of poverty, girls are easily perceived as a demand on household resources, conveniently transferable through marriage to the responsibility of a husband.Puberty rather than chronological age commonly marks the passage from childhood to adulthood.Girls receive instruction preparing them for marriage and sexual life in ceremonies signalling their availability for marriage (Munthali and Zulu 2007) and leading to the likelihood of early pregnancy.Outside marriage pregnancy is associated with shame, while early marriage is associated with a greater risk of domestic violence, separation, and divorce (UNICEF 2011).In such situations, young women's chances of securing independent means of survival through employment are limited to insecure and low paid jobs that are more commonly available in urban areas that draw them away from existing support networks.
The 2012-2016 Malawi Child Protection Strategy (UNICEF 2011) described the child protection situation for many Malawian children as being 'dire' with one in six children vulnerable to violence, abuse, exploitation, and neglect, and at risk both to and from HIV and AIDS.The foreword to Malawi's National Plan of Action for Vulnerable Children 2015-2019 (Government of Malawi 2015, p. 1) begins: "The child protection situation in the country remains dire.The situation analysis conducted in 2013 found that there are over 1.8 million vulnerable children in the country . . .many . . .subjected to various forms of abuse, exploitation and lack of access to essential services".Girls are reported as being slightly more likely to be vulnerable than boys (53/47per cent) with vulnerability conceptualised as not living with birth parents, living in households where no adult has had at least primary school education, or being an orphan of one or both parents.Among this group are an estimated 12,000 children living in child-headed households and around 10,000 in institutional care.While many terms referring directly or indirectly to children's engagement in sexual activity-early marriage, early sexual debut and teenage pregnancy, early sexual intercourse, sexual violence, and sexual abuse-are used in the Action Plan, the term prostitution does not appear.

Child Prostitution: Theoretical Perspectives
Prostitution has been defined by O'Connell Davidson (2007) as a term referring to the trade of sexual services for payment in cash or kind, a form of social interaction that is both sexual and economic.Children's involvement in prostitution, or 'child prostitution', as a social phenomenon has been conceptualised in different ways reflecting diverse and changing cultural understandings of childhood across the globe.In this section, we provide a brief critical analysis of three dominant discourses in this field: children's rights, feminist perspectives, and structure/agency debates.Arguing that all three perspectives are inherently adult centric, we explore the potential of the capability approach (Sen 1999(Sen , 2005;;Nussbaum 2005) to develop a more nuanced understanding that is generated by those with experience of involvement in prostitution as children.

Children's Rights
The children's rights perspective has been highly influential in debates about child prostitution for the last 25 years.Article 34 of the UNCRC (United Nations 1989) refers to children's right to be protected from all forms of sexual exploitation and sexual abuse including inducement or coercion to engage in any unlawful sexual activity, and exploitative use in prostitution.The subsequent Optional Protocol on the Sale of Children, Child Prostitution, and Child Pornography (United Nations General Assembly 2000), articulates the need to protect children against prostitution as a form of sexual abuse.The result is an equation between child prostitution and child sexual abuse (Pearce 2006;Brown and Barrett 2013).The children's rights perspective has moved perceptions of children that are involved in prostitution from being seen as delinquent (Phoenix 2002), to being seen as victims of child sexual exploitation in need of protection (Van Meeuwen et al. 1998;Hounmenou 2016;Montgomery 2001;Pearce 2014;Phoenix 2007).However, this has led to a generalised perception that children must have been forced into prostitution (Stacey 2009;ECPAT International 2011), supported by vivid imagery of living conditions characterised by abuse and exploitation, including being chained in brothels, drugged, raped/gang raped, and trafficked (Pearce et al. 2013).
Despite the impact of the children's rights perspective in recognising adult power over children it is limited by lack of attention to the diverse contexts of children's involvement in prostitution, depicting children as a single group, building on normative concepts of childhood.This has led to the marginalisation or exclusion of children that are involved in prostitution from critical engagement in decisions that affect their lives.It has perpetuated thinking that children who become involved in prostitution must have been subject to a process of 'being prostituted', and led to a narrow range of 'rescue' responses.Most studies of child prostitution rely on information from service providers, including police, and are produced by, or carried out on behalf of, organisations working with the aim of rescuing children from prostitution.This has led to stereotypical understandings of child prostitution (O'Connell Davidson 2005;Montgomery 2007;Orchard 2007).As Melrose (2010, p. 18) explains, this discourse does not provide a means to understand the often messy, complex, realities that underpin the lives of those who become involved.

Feminist Perspectives
Feminist perspectives have emerged in response to the moralist discourse in which prostitutes and prostitution are conceptualised as immoral and constituted as a social problem.The moralist view is often held in contexts of conservatism and strong religious beliefs, including Malawi, where prostitution was popularly understood to be a criminal offence until a recent High Court ruling (Centre for Human Rights Education, Advice and Assistance 2016).Feminist writers argue that the moralist discourse overlooks the roles of those who extort, exploit, control, and/or intimidate women in prostitution (Phoenix 2007, p. 79), but they remain largely unchallenged in law.
The feminist approach has generated two distinct responses: the radical and the liberal.The radical feminist perspective frames prostitution as male driven violence and exploitation of women (Phoenix 2012), based on the notion of power imbalances in paternalistic societies in which men control women who occupy powerless, peripheral positions with limited opportunities.Like the children's rights perspective, proponents of this feminist position stress that: "choosing sex work is impossible . . .sex work is per se a human rights abuse, and thus choice is negated, irrelevant" (Doezema 2010, p. 25).On this basis, radical feminists argue that prostitution should be prohibited, a view that is challenged by liberal feminist theorists and activists who promote the rights of sex workers on the basis that, precisely because society is structured paternalistically and women's opportunities are highly constrained, they may decide to engage in prostitution as a means of earning a living with the intention of overcoming their deprivation and constraints.In doing so, they are able to exercise agency and make rational, if constrained, decisions (Doezema 2010).In this sense, liberal feminism rejects the view that all women in prostitution are prostituted, forced against their will, and subjected to exploitation and abuse.However, it does acknowledge that sex work is a very risky route to livelihood and calls for measures to protect women from abuse and exploitation, decriminalising prostitution, and criminalising exploitative practices (Phoenix 2012).Limitations of the liberal feminist perspective relate to its lack of attention to social divisions within the broader category of gender, and in particular, its failure to work through the implications for children who are involved in prostitution and for male clients of female prostitutes.
In discussing the merits and limitations of children's rights and feminist discourses, Phoenix (2007) argues that prostitution can better be understood by complementing the acknowledgement of agency (rejecting arguments that rest solely on victimhood) with closer attention to historical, cultural, political, legal, and economic factors that shape the structural conditions of prostitution.

Structure and Agency
The importance of structure and agency as inextricably linked derives from Giddens (1986) theory of structuration, which conceptualises individuals as active agents, aware of, and negotiating with, different situations or structures.Just as social structures influence individual agency, with awareness of those structures, individual agency can influence social structures.Drawing attention to the importance of structural specificities and human agency, a structure/agency perspective has the potential to develop further understanding of (child) prostitution, focusing on how individual actors engage with particular structures to make particular decisions (Phoenix 2002;O'Connell Davidson 2005;Montgomery 2001;Orchard 2007).As Hwang and Bedford (2003) demonstrate, despite some similarities of circumstances surrounding children's involvement in prostitution, specific conditions, and motivations for joining, continuing, or leaving prostitution differ widely within and between countries.This demands ways of making sense of child prostitution that go beyond generalised overviews of what constitutes child prostitution, to access the understandings of those with experiences of involvement and the structural conditions in which they engage in prostitution.Limited studies adopting this approach (Montgomery 2001(Montgomery , 2007(Montgomery , 2014;;Phoenix 2001;O'Connell Davidson 2005;Orchard 2007) have focused largely on the contexts and structures in which children who are involved in prostitution make decisions and exercise agency.But, there has been little development in extending the use of this knowledge to explain why particular structures lead children and young people to make particular decisions, or to explore the impact of particular structures and particular decisions on children.It is here that we turn to the capability approach, expanding the framework of the structure and agency debate by examining the impact of structural factors, and the freedom that individuals have to make decisions that contribute to their well-being (Sen 1999;Nussbaum 2001Nussbaum , 2005)).

The Capability Approach
Underpinning contemporary measures of human development, the capability approach is centrally concerned with human rights and social justice, referring to the freedom of individuals to 'lead the kind of lives they value-and have reason to value' (Sen 1999, p. 18).The approach treats well-being as a matter of individual evaluation, rather than as something that is objectively measured as, for example, in Hounmenou (2016) structured survey of child prostitutes in Ouagadougou.The capability approach lent itself to the purpose of the study reported here, to develop more nuanced understandings of children and young people's involvement in prostitution based on their own experiences and their own ways of understanding their involvement.Nussbaum's role in developing the capability approach has focused thinking about its application in understanding women's and girls' human rights and capabilities.Unlike Sen, who argues that the determination of specific capabilities may vary depending on diverse cultural understandings of valued lives, Nussbaum (2003) has defined basic capabilities central to human life, including the ability to have: good health including reproductive health; adequate nourishment and shelter; security against violent assault; the ability to think and reason in a truly human way informed by an adequate education; emotional development not blighted by fear and anxiety; and, the ability to form a conception of the good life and to engage in critical reflection about the planning of one's life, the social bases of self-respect, non-humiliation, dignity, and worth equal to that of others without discrimination.Nussbaum (2005) outlines how structural and cultural factors can have a disabling effect on capabilities, especially those of girls and women in patriarchal societies.She argues that girls, not recognising viable alternatives, exercise adaptive preferences conforming to culturally normative expectations, and complying with structural inequalities.But, because the capability approach focuses on what individuals are able to do and to be, it challenges an over-emphasis on structural inequalities that give rise to exploitation rationalised on the basis of cultural relativism (Nussbaum 2005) and it opens a new space for making sense of children's involvement in prostitution.
Like the children's rights approach, the capability approach is thought by some to be heavily influenced by western thinking linked to individualist liberal ideas of what constitutes development and well-being (Jaggar 2006), paying little attention to power inequalities that are associated with structural factors (Dean 2009).However, the capability approach has been employed to explore questions of 'well-being deprivation of sexually exploited trafficked women' (Di Tommaso et al. 2009) and it has been discussed in relation to children in a range of settings, including education and the street (Biggeri and Comim 2011;Peleg 2014).The study reported here represents a significant development in the use of the capability approach as a framework for furthering understanding of children's involvement in prostitution, and how this can be understood as an issue of social justice beyond the logic of the children's rights perspective that equates sexual activity with children and sexual abuse.The approach also extends other perspectives and debates that inform understandings of child prostitution by exploring the freedoms and 'unfreedoms' of those who are involved to make choices that will enable them to achieve well-being.In this way, the capability approach stands out as an alternative lens through which to develop a more nuanced understanding of child prostitution.

Research Questions
Following this introduction to the politics and relevant policies in Malawi, and to the theoretical frameworks that have shaped thinking about child prostitution, we now move to the empirical study that addressed the following research questions: Why and how do (girl) children and young people in Malawi engage in prostitution?How do children and young people understand their involvement in prostitution?How do social, economic, political, and cultural issues influence and shape children and young people's involvement in, and routes out of, prostitution?

Design
The research questions called for a qualitative approach to generate rich, deep, contextualised, and nuanced multi-dimensional insights into the social phenomenon being examined in a specific context (Mason 2007).They also called for methods of data collection, analysis, and interpretation that would place children and/or young people with relevant experience at the centre of the process.
Given the legally and culturally sensitive nature of the research to understand children and young women's experiences of prostitution in a deeply patriarchal society, and the importance of emphasising dialogue in documenting participants' accounts, a participatory research framework was adopted (Kindon et al. 2007;Jupp-Kina 2010).Founded on principles of democratic participation, critical participatory research is acknowledged as an effective means of unearthing and articulating 'grossly uneven structural distributions of opportunities, resources, and dignity' of 'troubling ideological categories projected onto communities (delinquent, at risk, damaged, innocent, victim)' and 'contesting how "science" has been recruited to legitimate dominant policies and practices' (Torre et al. 2012, p. 171).Based on the belief that "human beings are dynamic agents capable of reflexivity and self-change" (Kindon et al. 2007, p. 13), participatory research facilitates the generation of knowledge through 'negotiated spaces' that are based on a commitment to sharing power so that the knowledge generated is of direct relevance and benefit to the participants involved (p.22).

Sample
Access to young people involved in prostitution was negotiated through two national NGOs whose work in the field of HIV/AIDS prevention included work with adult sex workers and plans to work with adolescents.The initial intention had been to focus on the experiences of this younger age group.However, with the NGOs being heavily dependent for funding on international donors that typically shape the focus of programmes and projects to reflect their own priorities (Morfit 2011), at the time of starting fieldwork the aspiration of both NGOs to work with adolescents had given way to broader HIV/AIDS prevention initiatives.This led to the recruitment of a purposive sample of young female sex workers who were involved in programmes to encourage safe sexual practices.While all but one of the participants were aged over 18, they had all engaged in prostitution as children.Adapting to the realities of participants' lives, and responding to what, in hindsight, can be seen as a serendipitous opportunity, the research was refocused to include an understanding of the connections between child and adult prostitution.The final sample consisted of ten women in their twenties, eight in their thirties, and one twelve year old girl, the daughter of a participant in her thirties.Having accompanied her mother to initial meetings she expressed strong enthusiasm to tell her own story.She was the only member of the sample who was not actively involved in prostitution at the time of the study having escaped from the brothel to which she had been lured under false pretenses and then forcibly confined.Only one participant had completed secondary education, and over half had not completed primary school.

Ethical Considerations
Scrupulous attention was paid to ethical considerations in this highly sensitive area of research in the context of an 'illicit economy' (Sanders 2006), and considering the (male) gender of the lead researcher whose role would necessarily involve direct engagement with girls and young women in a cultural context strongly influenced by traditional structures of male power and dominance (Kaime 2009).The process for gaining ethical approval, granted by Durham University Ethics Committee and Malawi's National Commission for Science and Technology's Committee on Research in the Social Sciences and Humanities, involved demonstrating robust arrangements for securing the informed consent of participants, guaranteeing their anonymity, achieved through the use of pseudonyms, anticipating risks, and identifying measures to avoid harm to all participants, including the researcher.However, mindful of the principle-based and regulatory focus of institutional ethics committees and calls for relationship-based approaches to ethics in participatory research (Banks et al. 2013), careful attention was also paid to the probability of meeting unanticipated challenges, and the need to be continuously mindful of 'situated ethics' (Perez 2017;Sultana 2007) in maintaining a feminist ethical stance (Usher 2012).The introductory stages of fieldwork were supported by two female research assistants who arranged meetings with participants and whose presence aided the development of trust and openness.A number of practical arrangements were made to ensure that contact between (female) participants and the (male) researcher would not risk compromising the safety of either (Sultana 2007).Meeting times, places, and spaces were arranged to avoid the male researcher being alone, or being perceived to be alone, with any participant and participants were able to contact a female research assistant, in confidence, at any time, should they have concerns or questions they felt unable to raise with the male researcher.Before consenting to participate, either in writing, or through audio recording, participants were assured they would be able to withdraw from the research, temporarily or permanently, at any time, for any reason or none, without repercussion.These measures to increase participants' confidence in the research process as being a shared endeavour enabled the research to follow the pace and daily realities of participants' lives, which was particularly important given their needs to engage with income generating opportunities that were neither regular nor entirely predictable.Recognising the economic circumstances of participants, they were compensated for their time, and, in recognition of their contributions to the research, with food for their families and drinks and snacks during meetings.

Data Collection
Early meetings involved discussions about the purpose of the research and how it could be shaped to maximise benefits for participants.As prostitution is highly stigmatised in Malawi, it was particularly important that participants should feel able to share their experiences without shame or embarrassment, and decide how they wished to contribute to the research.But, unused to being respected as girls and women in a patriarchal society, to exercising power in setting terms of involvement in family, educational, or working environments, and unfamiliar with taking part in research, participants asked for guidance about different ways they could share their experiences.A wide set of data collection techniques was presented and discussed with participants who showed interest in learning how different techniques could work in practice.The freedom to choose which techniques to use in order to convey their experiences enabled participants to express themselves in ways that were sensitive to their abilities, skills, and interests.They chose: (i) photography using digital cameras, following guidance on their use, to take photographs representing any aspect of their experiences of prostitution, followed by discussion of the significance and meaning of the photographs (Pain 2012); (ii) problem trees in which a drawing of a tree is used to identify problems requiring attention (represented by the tree trunk), their causes (roots), and effects (branches, leaves, and fruits), followed by the discussion of possible solutions (Chevalier and Buckles 2013); (iii) character drawings to generate representations of the self (Bagnoli 2009) or a fictional character that served as a device to talk about personal experiences of prostitution as if speaking about a third person; (iv) storyboarding (Thomson 2009) that, like character drawing, facilitated telling stories of involvement in prostitution in the third person, but differed by focussing on sequences of events marking the course of involvement in prostitution; and, (v) paper cutting and ranking (Chevalier and Buckles 2013) to record information they did not wish to speak of directly and openly in a group but were willing to share anonymously on paper (with those unable to write aided by a trusted friend) to stimulate collective discussion.Examples of issues discussed in this way included age of first engagement in prostitution and priority concerns about experiences of prostitution.Narrative interviews were conducted with each participant to elicit the meanings and significance they attached to their images, enriching the data and addressing emerging gaps through continued discussions (Ansell et al. 2012).
Special care was taken in relation to data collection with Loni, the youngest participant, recognising her age-related vulnerability.Four meetings were held with her at a centre frequented by participants, with her mother and another trusted participant nearby, but out of ear shot.The first meeting was to ensure that she had sufficient information and understanding to give her own fully informed consent.In the second, she was introduced to different possible methods to convey her experiences.The third and fourth meetings centred on her construction and discussion of a character drawing through which she conveyed her story of involvement in prostitution.
Free to choose their preferred techniques, the flexibility of participatory research (Robson 2011) came to the fore.Multiple data sources captured participants' experiences and knowledges in ways that they had chosen.Taking advantage of opportunities to engage in activities that were new to them was perceived by participants as a benefit of the research.The result was the production of rich data that were discussed further, in groups and/or individually nurturing collective and self-reflection without becoming hostage to methodological claims of validity and reliability (Kemmis et al. 2013, p. 70).
Two further, semi-structured, interviews were conducted with representatives of government social welfare departments operating in the research sites in order to understand the role of state social welfare structures in working with children engaged in prostitution.

Data Analysis
Nineteen participants generated a rich and diverse data set consisting of 194 photographs ranging from places where they met and engaged with clients to representations of what life might be like if it were not necessary to engage in prostitution, nine problem trees, 18 character drawings, two storyboards, five sets of paper-cuttings/rankings, and 96 audio recordings from discussions and narrative interviews that were conducted in Chichewa.Data analysis was an ongoing process starting with the identification of emerging themes during participants' discussions of the data they had generated.Audio data were transcribed in Chichewa to retain accuracy and originality, followed by translation into English, and NVivo was used to assist with the management of this large quantity of data.This was supplemented by the lead researcher's reflections in making connections, and identifying gaps, between participants' experiences and existing theoretical explanations for childhood prostitution.The development of a 'mind map' linking themes and interpretations (Wheeldon and Faubert 2009) revealed a dominant picture of journies into prostitution, experiences of prostitution, and thoughts about future life prospects.

Results and Discussion
Here, we present and discuss the results of this empirical study by addressing each research question in turn.

Why and How Do (Girl) Children and Young People in Malawi Engage in Prostitution?
Nearly all of the participants had first engaged in sex in exchange for money or goods as children.They used expressions including ali mwana (child), wachichepele, wang'ono (young/younger), and ali ku primary (while at primary school) when talking about their own initial involvement and that of children and young people currently joining prostitution.
Dominant and interconnected themes underpinning participants' stories were the absence of family support, whether through extreme poverty or orphanhood, and the role of cultural norms that proscribe early marriage for girls, limiting opportunities for educational attainment and independent income generation.An additional factor for some was their resistance to normalised gender-based violence within marriage.But, having left their marriages, they found that prostitution was the most feasible way of generating sufficient income to meet survival needs.A small minority, however, explained that they had first engaged in prostitution after being introduced to transactional sex (Kaufman and Stavrou 2004) at the suggestion of friends or relatives who were already engaged.The youngest participant's story vividly demonstrated her experience of coercion into prostitution, while several other participants described practices of coercion and trafficking that they reported as being more common amongst younger girls.
Poverty was the principal reason recognised by participants for engaging in prostitution, perceived as the only feasible means of meeting their own basic needs and often those of their children, siblings or parents.Despite the urban locations of the fieldwork, on the basis of their own earlier experiences and knowledge of other girls arriving in urban areas, participants commented on the similarity of experiences for girls whether they lived in rural or urban areas.Bureni, who first exchanged sex for money as an adolescent, explained: Mostly, it's because of problems; being without food because my parents were destitute . . .I would buy food and support my mother with the money I earn from prostitution.Asked why she felt it was her responsibility to provide this support she replied: Because she could not afford food.She lacked support.I would then buy clothes for her.
Orphanhood, associated in Malawi with the loss of one or both parents, was a common experience among participants.With husbands and fathers being widely perceived in Malawian society as responsible for generating the means for family survival, and with little or no protection in terms of inheritance rights, the loss of a father through death or divorce is associated with loss of the primary source of family income.Mpaseni, for example, first engaged in prostitution following the death of her father.At that time she was still at school.She explained that her father's relatives: took everything.They left us without anything.They also collected my father's pension on pretence they were looking after us and paying for our school fees.Her mother, unable to support Mpaseni and her brother, returned to her home village.But still unable to meet their basic needs Mpaseni accompanied her mother to Lilongwe, the capital, on a daily basis, to beg.Mpaseni subsequently married a street trader and was briefly able to stop begging, but following her own husband's premature death she was once again deprived of a culturally normative means of survival and turned to prostitution.
Orphanhood was no less challenging for those whose mother had died.As Labani explained: losing your parents can drive you to prostitution particularly if it is your mother because your father will find another woman who may be abusive or not care enough for you . . .this is likely to make you consider other options rather than embrace the mistreatment.She went on to explain: what else could I do except this?But that does not mean I like what I am currently doing.I have tried to leave before . . .I once went to South Africa as I was trying to find things I could do.But even there I was working like a slave [so] I returned home and rejoined sex work.
A similar theme emerged from Diami's story.She explained that she and her sisters had been taken in by their grandparents following the death of both parents.But, they had been overworked and deprived of food.Their response had been to run away which, in the cultural context, was seen as an act of rebellion, yet one that can also be interpreted as one of resistance and resilience (Williams 2010).Wider group discussions revealed a common understanding that extended family structures could not necessarily provide a safe environment for children who had been orphaned.As Babani explained: There are some orphans who manage to be fostered and taken into care by very supporting and loving relatives.Maybe we were just unlucky that our parents' relatives were cruel.For the participants in this study, prostitution offered a means of meeting daily needs in response to poverty, exacerbated by violence that is associated with a strongly patriarchal society and a lack of public services to support families in meeting basic survival needs, even when assuming responsibility for orphans.However, this knowledge alone does not explain how they become involved in prostitution.
Participants were aware that not all girls and young women in similar circumstances became involved in prostitution, and their stories included strong clues about additional factors that pulled and pushed them into prostitution.They drew attention to the roles of family members and friends in their decisions to engage in prostitution.Some considered involvement as almost hereditary, or a family occupation, for example, Bureni, who explained: My cousin who is currently in Lilongwe started when she was young.She has even been to Mozambique where she had followed her mother who is also a sex worker.Mpalini added another dimension to Bureni's explanation saying: Bureni used to admire those of us who were doing sex work.Much as we shared the food we prepared from our earnings with the rest of the family, we always got the lion's share to go with the labour pain of our involvement.
Group discussions concluded that growing up in a household or neighbourhood where relatives and close friends were selling sex would enable children to normalise prostitution as a way of meeting their needs.But importantly, participants stressed that it was only in circumstances of poverty and hardship that the influence of social networks would tilt children and young people into perceiving prostitution as a possible form of livelihood.They also drew connections between poverty, early marriage, and withdrawal from school.While cultural expectations of marriage in adolescence for girls divest parents of financial responsibility for daughters, participants also associated early marriage with risks of domestic violence, pre-mature pregnancy, early withdrawal from school and denial of opportunities for economic advancement.For example, Mazani presented a photograph of a car and her understanding that the female owner must have: reached the point of buying the car because of school.Referring to the wider discussion Labani explained: We pondered on what could have happened if we had also gone further with school ... if only we had had all the support we needed, we could have reached her stage.
The route from poverty to prostitution was not direct.Exercising ambiguous agency (Bordonaro and Payne 2012) some participants had reluctantly accepted offers of poorly paid employment as house maids or bar workers where they found themselves experiencing abuse and exploitation.Nussbaum (2000Nussbaum ( , 2003) ) refers to this process as the exercise of constrained choices or adaptive preferences.In hindsight these offers of employment were understood by participants as thinly disguised inducements to engage in prostitution.As Mpaseni explained: There are some girls who are deceived with promises that they will be working as maids or ushers at a restaurant only to be told that they will be working in prostitution.The pub owner tells them that he will be collecting money from the clients and that they would be paid on monthly basis.Twelve year old Loni described her own experience: We were not informed in advance of the nature of the work or how much we would be getting.We were only told upon arrival to start opening beer for people.I was very shy and nervous.But they said I will not be getting food if I insisted on not working.Once in the grip of brothel owners young prostitutes were controlled through threats of non-payment, violence and food deprivation.Loni continued: I was afraid of the owner who his friends called Soldier.Some said he was given a 30-year jail sentence for murder but he managed to break out of prison.I thus would just submit to whatever he said.Kept out of sight of any client or visitor suspected of being a police officer, Loni was prevented from attracting the attention of authority figures who she hoped, recognising her extreme youth, might have helped her.

Understanding (Girl) Child Prostitution-Participants' Perspectives
Explaining why and how they became involved in prostitution, it was clear that participants did not see themselves simply as passive victims.At different points on their journeys they had made active decisions that led them directly or indirectly into prostitution as a means of assuring their own, their children's, siblings', or parents' survival.But once involved, they found themselves exposed to various forms of exploitation and abuse.Despite recognition of the roles of poverty, early marriage, and withdrawal from school as drivers of childhood prostitution they did not articulate their understandings in terms of the broader structural factors behind these circumstances, leading to their involvement in prostitution.
With the exception of Loni, aged 12, participants viewed their involvement in prostitution as 'kuutsanza mtima' (personal choice depending on circumstances).But, they also identified with experiences of exploitation and victimhood.A vivid example was offered in a discussion about the hardships endured as a prostitute that arose from Emoni and Diami's storyboards depicting non-payment by clients, refusal to use condoms, violent rape, and threats of murder.Yet, Jobani used the term 'iweyo wamuyamba mwamunayo' (clients cannot be violent or abusive unless you instigate it) seemingly promoting the benefits of women's submission, a view reflecting feminist understandings of the marginal positions held by women in patriarchal societies.
While engaging in prostitution was an illustration of participants' agency in attempting to better their lives, it is also an illustration of ambiguous agency (Bordonaro and Payne 2012), and adaptive preferences (Nussbaum 2001).Likening their experiences of clients' abusive behaviours to those encountered in marriage demonstrated awareness of male hegemony across social institutions.The value of the capability approach here lies in offering a more nuanced understanding of the circumstances and motives behind the decisions to engage in prostitution as well as the effects of those decisions on the well-being of participants.This is particularly important in identifying appropriate ways of addressing the complex combinations of circumstances that draw girls into prostitution.
Despite experiences of abuse and exploitation participants had chosen to continue their involvement since, despite threats to a number of central capabilities (Nussbaum 2001) it seemed to offer the only way of assuring survival.As Bureni put it: no-one quits prostitution.Other participants described themselves as 'chiwanda cha uhule' (possessed by prostitution).Rather than providing a means of livelihood, prostitution involved facing challenges of non-payment, violence and coercive control.But, the lack of alternative means of survival left them 'choosing' to continue their involvement.
Loni's story, an 'information rich extreme case' (Palinkas et al. 2015), powerfully demonstrates the importance of avoiding unidimensional theoretical explanations for children's involvement in prostitution.Loni recognised her involvement as constituting abuse and exploitation having been tricked under the pretence of employment as a maid before being coerced into prostitution.But, her initial search for employment demonstrates her agency in the face of social structures that threatened her survival, and in her quest to achieve a life she could value.Once she had been coerced into prostitution she became increasingly controlled by the brothel owner pressurising her to meet the sexual demands of clients.Her own attempts to explain to clients that she was not old enough to have sex were unsuccessful and met with physical attacks and injuries that were sustained through clients' attempts to have penetrative sex.In this sense, Loni did see herself as a victim and her experience of prostitution as abusive.She used the expression 'anabedwa' (stolen against her will), speaking to the children's rights approach to child prostitution.She also used the term 'mwana dala'3 referring to her perceived value as a sexual object precisely because of her young age, a perspective that speaks to the radical feminist discourse (Phoenix 2012;Doezema 2010).
Despite clear evidence of abuse, Loni's story demonstrates agency throughout.One of five children of a lone mother who was also involved in prostitution, Loni had sought ways to help support her family.Her age and lack of educational achievement meant that the only employment opportunities open to her were in the informal economy where wages are low.Her search brought her into contact with traffickers who tricked her into accepting a job offer that, in reality, was a thin disguise for prostitution.Isolated from her family, with her earnings controlled, threatened by non-payment and food deprivation, Loni 'conformed' to the circumstances in which she found herself before exercising agency in the strongest way by planning and achieving her escape.She returned to her mother's home and involved the police who intervened to 'rescue' others being held against their will in the brothel.Despite her experiences, her original motivation to support her family led her to reflect: 'sindikanapangira mwina' (I could not do otherwise).

Protection of (Girl) Children Involved in Prostitution
We turn now to consider the question of the protection of girl children engaging in prostitution, framing our consideration from an inequalities perspective.
Reflecting the lack of explicit recognition of child prostitution in Malawi's National Plan of Action for Vulnerable Children 2015-2019, interviews with two district social welfare officers in the fieldwork sites confirmed a lack of specific policies or programmes to prevent or respond to child prostitution.Instead, children who were involved in prostitution were assimilated into the broader category of street children, remaining largely invisible in programmes to combat the devastating impact of HIV/AIDS on the livelihoods of Malawian citizens.This contrasts with other categories of 'vulnerable' children, including child labourers on tea and tobacco estates and girls experiencing early marriage, and it arguably reflects attitudes surrounding premarital and extramarital sex as symbols of masculine prowess, while sexual activity outside of marriage for girls is strongly discouraged and stigmatised (Kelly et al. 2013).State responsibility for child protection lies with the Ministry of Gender, Children, Disability, and Social Welfare.However, state capacity to address issues of child protection are limited both by national poverty and the persistence of a patriarchal society that slows the rate at which legislation to address a range of inequalities affecting children, and particularly girl children, can have a positive impact.Furthermore, much financial and capacity building support, including the development of social work training, is provided by international donors and organisations, implying the use of western models that are questionable in the Malawian context (Kakowa 2016).
Malawi's position on the Human Development Index, reflecting severe problems in the fields of health, education, income, and gender inequality, leaves many children without an effective means of support (Kidman and Heymann 2016).Under these circumstances, together with the stigma surrounding sex outside marriage for girls and women, it seems likely that explicit attention to child prostitution has been viewed as an unachievable policy goal.However, we argue that an understanding of inequalities in the protection of child prostitutes is also rooted in the stark inequalities that are associated with geopolitics further limiting state capacity to address social problems.These include impoverishment that is associated with tax avoidance by multi-national corporations that limit the potential for tax revenues (Dahlbeck 2016;Conroy et al. 2006); austerity conditions accompanying international interventions to support economic development that limit investment in public services; and, programmes funded through bi-lateral aid and international NGOs that are heavily influenced by values of neo-liberalism that fail to account for complex cultural specificities and are limited in their effectiveness (De and Becker 2015;Özler 2016).The failure or limited impact of these interventions is also commonly associated with poor internal governance and allegations of corruption.While more nuanced than commonly understood (Anders 2010;De Maria 2010) these problems are acknowledged as having played a part in limiting the effectiveness of poverty reduction plans and programmes.

Implications
Here, we reflect on the insights from the empirical study that, despite the limitations of scale and specific location, reveal clear implications for theoretical understanding, methodological opportunities, policy and practice challenges, and future research in the field of child protection and social inequality.

Theoretical Understanding
Reflecting on different theoretical perspectives of child prostitution, the data generated in this study revealed vivid examples of the violation of children's right to be protected from all forms of sexual exploitation and sexual abuse, and exploitative use in prostitution.Also evident were the roles of structure and agency, and of both radical and liberal feminist approaches that characterise prostitution, respectively, as exploitation and commodification of women's bodies, and freedom of choice within a wider context of gender inequality.Without explicit reference to the basis of exploitation, the participants' stories were full of references to age and gender that illuminated the complexities and cultural contradictions of prostitution originating in childhood, and that cut across the children's rights, radical and liberal feminist, and structure/agency theoretical frameworks.The hegemonic power of men was strongly evident in their stories in ways that demonstrated an acceptance, albeit unconscious, of existing social structures.In common with Phoenix (2001) study of how women make sense of their involvement in prostitution in Britain, Orchard (2007) study of child prostitution among young Devadasis in India, Montgomery (2014) discussion of child prostitution in Thailand, and O'Connell Davidson (2005) research on children in the global sex trade, this study of child prostitution in Malawi takes the understanding of child prostitution beyond competing theoretical claims, connecting explanations from different theoretical perspectives to capture the complex nature of girl children's and young women's experiences of prostitution.Using the lens of the capability approach, the impact of structural factors on girls' and young women's well-being in the cultural context of Malawi became clearer, as participants demonstrated the ambiguous nature of their agency and the adaptive preferences that they made in a context of constrained choices.These insights are particularly helpful in understanding why attempts to 'rescue' girls and young women from prostitution are often unsuccessful (Rafferty 2016) and help to make sense of Bureni's claim that 'no one quits prostitution'.

Methodological Implications
Although the study has generated valuable insights into participants' understandings of their involvement in prostitution, in childhood and young adulthood, claims for the value of participatory approaches must be treated with caution.Referring to the 'tyranny' of participation, Cooke and Kothari (2001) have pointed to the ever present dangers of falling short of the principles of participation by failing to achieve power sharing over decision-making in research processes, failing to recognise power differentials among participants that can lead to the reinforcement of existing power structures, and failing to consider that participatory approaches might not be the only strategies to investigate particular problems.Conscious of these concerns, care was taken to ensure that participants were fully engaged in decisions about the research process, leading data collection activities and sharing in analysis and interpretation to ensure that knowledge was generated through a process of on-going negotiation.They shared in dissemination of findings to the NGOs and joined a team of sex workers in successfully challenging the legality of the criminalisation of prostitution, supported by funding from a local NGO.
Reflecting on alternative approaches to researching the lives of children in prostitution there is significant debate about the validity and reliability of findings from participatory studies, particularly in cultural contexts where the stakes of speaking honestly about involvement in premarital or extra marital sex are high.Those favouring the use of surveys argue that the presence of human interviewers is likely to introduce bias, though Poulin (2010) study comparing the use of a standard survey and an in-depth interview with 90 never-married young adults in rural Malawi to gain information about first sexual encounters, demonstrated greater honesty in the interviews than the survey.Poulin argues that the flexibility and reciprocal exchange of the interviews fostered greater trust and more truthful reporting.
We argue that a power sharing approach embracing participants' own choice of methods, and continuing dialogue to support the expression of experiences and understandings of their involvement in prostitution, has delivered credible insights that would have been difficult to access while using more conventional research designs and methods, either quantitative or qualitative.We make no claims to the representativeness of the findings, since they refer to experiences in a particular cultural setting at a particular time in political history.But reflecting on debates about the validity of participatory research findings, we argue that the study meets the criteria for methodological rigor, relevance and reach outlined by Balazs and Morello-Frosch (2013).Following Van der Riet (2008) argument, by focussing on the relational and social nature of human action, the participatory processes have taken us beyond the 'moral imperative' and social justice motivations that typify much participatory research, to enhance validity by illuminating understandings and interpretations of human action.Offering feedback on the use of different visual methods, one participant, Mpaseni, referring to a particular drawing explained: the process of drawing enabled me to look back and speak about my involvement in prostitution from the time I started, while another, Zani, said: We are able to go deeper and vividly express ourselves which could have been very unlikely to achieve as we could have felt uncomfortable to talk.Each person is able to talk about her story and experiences.Their feedback contributes a helpful 'participant perspective' on the validity of the data.

Implications for Policy and Practice
We argue here that efforts to support girls and young women to avoid or leave prostitution that disregard the strength of cultural practices that commodify girls and young women as housewives and child-bearers in waiting are unlikely to be effective.Perceived as a threat, by wider family and community interests, to normative expectations for girls and young women (Groves and Hinton 2013) such efforts easily meet with resistance that is poorly understood.Policies to tackle chronic gender inequalities in child protection must have resonance for both national and international political actors, acknowledging similarly chronic inequalities between countries of the global north that dictate the terms of international aid and citizens of the global south who may exercise the power that they have in resisting externally imposed solutions to social problems.
Interventions to address what is perceived as the injustice of having to resort to prostitution to survive, must be informed by a deep understanding of the cultural as well as economic factors that lead to the normalisation of transactional sex, and the broader impact of children's involvement in prostitution on society as a whole.International donors and local service providers must understand that social change involves challenging complex power relationships requiring engagement not only with national policy makers, but with grass roots communities, critically considering the congruence or incongruence between political, organisational, and community values and practices.This implies the identification of complex and interconnected threats to livelihoods, ensuring that legislation to address gender inequalities and gender-based violence is robust, fit for purpose and free of contradictions.It also implies finding effective ways of demonstrating the long term benefits of additional years in education for both boys and girls, whilst also addressing the short term challenges for families that this implies.
The use of the capability approach to underpin the Human Development Index suggests that it could realistically be adopted by policy makers to offer a multidimensional way of conceptualising child prostitution, drawing attention to structural factors, as well as individual behaviour and experiences, and linking child prostitution to questions of national development and social justice.Paying attention to structural factors that diminish the freedoms and choices of children and young women in Malawi-a patriarchal society that lends support to the continuation of early marriage and childbirth, and acceptance of girls' limited engagement in formal education-would illuminate more clearly the discriminatory experiences of girls and young women who are involved in prostitution.This, in turn, suggests closer attention to and engagement with children and young people involved in prostitution.While most NGOs involved in this area of work are informed by goals of HIV/AIDS prevention or control, children and young people are largely ignored as beneficiaries of such projects that tend to be targeted at adults.But as the study reported here has shown, the connection between child and adult prostitution is strong.While adults who are involved in prostitution were able to identify efforts of organisations to support them, they were clear that this was not the case while they were children.
Operationalising the capability approach for practical use in driving change for social justice presents significant challenges.While Sen declined to specify capabilities arguing that diverse cultural understandings of valued lives implied that capabilities should be determined locally, Nussbaum's list of basic capabilities that are central to human life is unlikely to gain legitimacy in all contexts.Robeyns (2005) argues that good practice in research for developing capability lists requires agreement about the nature and scope of the capabilities to be adopted, and agreement about the procedure by which the list of capabilities is to be agreed.This points clearly to the challenges of achieving agreement in contexts where deeply embedded cultural values have created dominant ways of thinking that, privileging powerful groups, are accepted as the norm.This effectively curtails the freedom and opportunities for the development of deliberative democracy required to develop an agreed list of capabilities.An example from the study presented here was the participants' views of sexual violence as something normal, to be expected, and even caused by them.But, despite these challenges, we argue that steps towards the articulation of valued capabilities can be developed among groups (Ibrahim 2006), in this case, those with experience of involvement in prostitution as children.Such opportunities for collective thinking and action can be stimulated by individuals and collectives, and supported locally, nationally, and internationally, to facilitate the recognition of agency and opportunities for exercising that agency with fewer or weaker constraints.Such a process was seen in this study when participants joined forces with other sex workers to challenge the illegitimate treatment of prostitutes as criminals.

Opportunities for Further Research
Discussion of policy implications arising from this study refer to the need for greater understanding of the factors that slow or prevent change not only to protect children more effectively, but to do so more equitably.Here, we make brief suggestions for further research opening up spaces to extend understanding of the effectiveness of interventions to enable children to avoid involvement in prostitution, to adopt the safest possible practices while they are involved in prostitution, and to support them in finding alternative routes to live lives that they value.These include: (i) further use of the capability approach to frame research for and by child prostitutes and those at risk of prostitution, (ii) longitudinal research to develop a clearer understanding of the longer term outcomes of engagement in child prostitution, (iii) an evaluation of the effectiveness of the Marriage, Divorce and Family Relations Act (Marriage Act) 2015 intended to protect girls from early marriage, thereby reducing early pregnancy and school drop-out among girls, and (iv) research to develop clearer understandings of the demand side of (child) prostitution offering insights into the potential for influencing the behaviour of men through challenging the logic and values of long established patriarchy and associated gender-based violence, what Jones and Jemmott (2016) refer to as cultures of male impunity and entitlement.

Author Contributions :
This article draws on doctoral research conceptualized and conducted by P.N. supervised by H.C. The article was drafted by H.C. and revised by P.N. and H.C. Funding: The study was funded through a Durham University Doctoral Studentship.