Performing the border of child labour: Roma working children

Abstract

The article explores the social, cultural and economical processes that lead Roma children into labour and their own interpretations of the value and risks of working. Based on qualitative research in several Roma communities in Romania the article analyses the different family strategies in coping with the economic difficulties of transition and the place of children in this process. The article is interested in the relationship between children and family, school and community and attempts to decipher what aspects in these relations encourage early entry into work. It argues that Roma children do occasional, poorly skilled, work that is relevant in their family economy, but invisible in and acceptable for broader society. Ultimately, the article argues that casting the situation of Roma working children in the framework of the child’s right to work carries the risk of reproducing racial and class prejudices.

1. Conceptual puzzle

Child labour is notoriously difficult to define and (arguably) culturally bounded. It was discovered as a “social problem” in Victorian England and was regarded as an unquestionable matter of poverty or cultural backwardness, until recently, when children’s relations with work became more significantly a matter of controversy.

There is, on the one hand, the legislative approach that tries to regulate children’s involvement in work that is abusive, exploitative and potentially harmful for overall development and education. The International Labour Organization (ILO) formulated labour standards and conventions that stipulate the acceptable and unacceptable work according to children’s age, workload, working hours and recreation time, (social and physical) environment, working rights, access to health and education.

ILO has developed categories to distinguish between different forms of work. Light work is defined as children’s participation in economic activity that does not negatively affect their health, development or interfere with their education, and which can be positive. Child labour is defined as children below 12 years of age working in any economic sphere, those aged 12 to 14 who are engaged in harmful work, and all children engaged in the worst forms of child labour and the worst forms of child labour, which refers to children being enslaved, forcibly recruited, made to enter prostitution, trafficked, forced to do illegal activities and exposed to hazardous work (ILO 138/1973).

It is only recently that the more invisible, but no less damaging, situation of children (especially girls) undertaking domestic work, came to light. Several risks that have not previously been considered refer to physical and psychological burnout, accidents, (sexual) abuse and neglect of schooling. Children’s work in agriculture also appears to be more hazardous than it used to be because of the increased exposure to chemicals and the global demand for cheap labour.

On the other hand, sociological and ethnographic research challenged the ILO position. The difference between work and labour, between harmful and non-harmful, between acceptable and unacceptable has become rather grey areas that attracted cultural dilemmas (White 1999; UNICEF 1997, cf. Liebel 2004: 195). It is argued that ILO policies were shaped by the Western model of childhood.

Whereas ILO generated a tendency to victimise working children, to protect, rescue and rehabilitate them from involvement in work that was criminalised, a new perspective has been given voice lately. First, an increasing number of children from the developing world ask for their right to work in decent conditions¹ and challenge the ILO recommendations that appear to respond to a Western notion of childhood. Besides, many children from the developed world are also working: not for ensuring survival and without economic pressures (Liebel 2004; Morrow, cf. Mayall 1994). It is in these circumstances that children start entering the academic agenda as actors, able to understand and improve the conditions they are living in and to voice their own concerns.

In the final analysis, all the above approaches have their advantages and limitations. ILO classifications, although necessary in order to prioritise interventions, were charged for being too coarse and theoretically implausible (White 1996, 1999; James et al. 1998; Liebel 2001, cf. Lieten 2004: 46). The agency-focused perspective, although put at its centre children’s own views on work, may risk overlooking the harsh conditions and exploitation that brings many children into work.

An alternative proposed was to regard children’s work as a continuum, from the light and positive forms to the most dangerous types (White 1996, 1999). Another, more radical, idea was to consider “any child out of school” as a potential labourer, because sooner or later he or she will be involved in a certain type of work (MV Foundation, India). Although this last definition risks oversimplifying children’s work, it may offer a practical starting point in considering Roma children who are working in Romania, given that statistics on education are more consistent than the ones on work.

Following Liebel (2004) and Boyden et al. (1998), this article will use the terms “children’s work” and “working children” which do not carry an implicit set of assumptions. “Child labour” will be used only when it appears there are obvious reasons that may render the work detrimental and harmful, or when citing ILO sources.

2. Working children in Romania

Far from being an apolitical issue, in Romania and in many other East European countries, child labour is rarely addressed. Political agendas are often attached to the topic, and its existence is often concealed or “manipulated”. There is no common term to differentiate between “labour” and “work” in the Romanian language and “child labour” seems far from receiving social recognition. It is occasionally presented in the media as a social problem when it is invariably linked to poverty. The middle-class practice of children’s part time employment, so prevalent in the West, is just emerging. However, in spite of a relative lack of awareness, children from Roma and majority population do work in Romania, whose total population under 18 years of age is approximately 4 million.

Statistics on the incidence of child labour are debatable, ranging from 70,690 (according to the National Institute of Statistics 2003) to 300,000 (according to individual economic estimates of C. Ghinăraru (2004) and to the International Confederation of Free Trade Unions (2005). The official statistics allow for limited international comparison, as the definitions used, differ to some extent from international ones. As no ethnic-based data is provided on child labour, there are no statistical arguments available to argue for a distinct “phenomenon” of “Roma child labour”. It is only by corroborating other statistical indicators that the focus on Roma ethnicity is gaining more legitimacy.

According to the National Census (2002), there are 535,140 Roma in Romania, accounting for 2.5 per cent of the total population. However, the number is prone to various biases, as Roma group membership is defined by self-identification, while ethnic identity itself is contextual, fluid and hard to define. What is more important are the generational dynamics inside the Roma minority. Whereas for the general population, 19.2 per cent are children under 14, for the Roma more than 34 per cent is made up of children below this age (CASPIS 2002).

What Roma children actually do matters within the inner structures of their communities and more generally in the dynamics of Romanian society as a whole. At present despite an increasing tendency for school attendance, Roma enrolment in primary and secondary education is still 25 per cent, and is 30 per cent lower than the national rate (ILO 2005). 17.3 per cent of Roma children aged between 7 and 16 have never had any form of education (ILO 2005). According to government sources² 80 per cent of children from the Roma population live in poverty, out of which 43.3 per cent live in conditions of severe poverty. The subsistence economy, unemployment and discrimination have excluded many Roma families. Child labour has been a survival strategy for many of them, an alternative to school failure for others, and a deliberate choice when a “promise of success through education” went astray.

Despite the fact that a child out of school risks being drawn into the world of work and that many working children struggle to comply with school requirements, education and child labour are regarded mostly as separate issues in Romania. In the literature except for a few studies (Surdu 1998, 2002; Cace et al. 2002) Roma children are invisible members of their communities. There are also very few ethnographies of any Roma group in Romania.

3. Research questions

The article aims to explore the social, cultural and economical processes that lead Roma children to work, children’s own interpretations of the work they are undertaking, and its perceived value and risks. It will analyse different family strategies in coping with economical difficulties of transition and the place of children in this process. The paper is interested in children’s relations with family, school and community in an attempt to decipher what aspects in these relations is encouraging early entry into work.

Is poverty the only given explanation or there are other mechanisms that influence Roma children’s work? Is the conventional belief that labour is a useful preparation for life a relevant explanation? Ultimately, what role(s) does Roma ethnicity playing in the broader process of early entry into work?

4. Methodology and limitations of the study

I was interested in documenting the experiences of Roma children and parents’ understanding of work. The topic itself is sensitive and can generate value-driven positions: from protectionism and victimisation, to an understanding of work as an expression of children’s autonomy. I tried not to privilege any scholarly position circulated in the literature.

Data collection and data analysis were simultaneous processes. The research did not start with a predefined sample, as is customary in sociological surveys on child labour. Especially when the purpose is to reach the “hard to reach”, this might have risked reproducing the exclusion of groups that were less visible in previous studies (given that the main classifications are made according to the “type of work” and children undertaking atypical work may have risked being overlooked).

The unit of analysis was not the individual (as in quantitative studies) but the event (e.g. occasional school absenteeism due to work; domestic work in one’s own household; children’s pride in work). The collection of data was a process that progressively sought to include experiences (events) that differed from those already collected (e.g. combination of school with work; domestic work in households other than one’s own; children’s feelings of shame and stigma). This strategy enabled the collection of rich data.

Nevertheless, by interviewing children and their parents from different communities (in regard to urban/rural environment, level of development, level of traditionalism, isolation from more affluent areas), more working experiences emerged. In some cases, the drawback was a reduced concentration on the community dynamic in which children’s work was embedded. Still, as there appears to be only one sociological study interested in children’s perspectives on their work in a Roma community from Romania (Cace et al. 2002), this article responds to the need to put forward Roma children’s work as a broader practice.

The research did not aim at statistical representativeness. Instead, it was interested in the theoretical or problem of representativeness (Strauss and Corbin 1990). Its final outcome consists in types of working experience and children’s responses to them. The research was explorative, looking into the diversity of processes, meanings and responses that characterise the work of Roma children. It was interested in the experiences of a “population” that may escape the standardised data collection procedures (based on questionnaires, with predefined work categories to be completed) and it aimed to challenge the assumption that Roma children are a homogenous social group.

The research was based on semi-structured interviews with Roma children and parents from six communities. It used participative and non-participative observation and the use of drawings for facilitating data collection (younger and less communicative children were invited to draw any of their working experiences, which mediated the relationship between the adult researcher and the children and evened out the power balance).

A large part of the previous studies on children’s work tended to collect data by questionnaires, or to gather the information from adults (Liebel 2004: 48). However, when applied to children, written methods (be they questionnaires, journals or essays) tend to produce socially desirable data and above all, to exclude children who are less literate but who have working experiences to share. Besides, we need to acknowledge that too much power is already embedded in a research relationship with children (measured by age, class, ethnicity, etc.), to ask children to complete a questionnaire. Open procedures that allow children space to develop their own ideas, help make the move from understanding children as objects to seeing them as subjects of study (Boyden 1998; Lieten 2001; see Liebel 2004: 48).

The interviewed children were between 9 and 17 years old. Fieldwork took place between June and September 2006 and 2007 and in December 2007. The children belonged to rural and urban communities from Transylvania and the South of the country. Several children were interviewed more than once and at different times (one year between interviews). The topics covered in interviews evolved during the research, depending on the “events” that were generated and which needed to be explored further.

The great majority of the interviews were recorded and transcribed. QSR Nvivo7 was used for facilitating data analysis. First-level categories were generated and they determined the need for further collection of data. The software facilitated the writing of Memos and the creation of relations between categories that were generated, based on successive comparisons. The communities that informed the fieldwork were:³

– F., a suburban community from Transylvania, with children working as daily “employees” at a poultry farm, collecting recycled metals from the city nearby or extracting and selling sand from a nearby river.

– A group of Rudari⁴ (Engl. Boyash), living near the Danube. Interviewed Rudari identify themselves as being a distant branch of Roma, “in between” Roma and majority population.⁵ They form a closed community, with a high degree of control exerted among the members, despite a high degree of mobility inside the country and, more recently, abroad (Spain). Rudari do not speak Romani language. This group’s main occupation is the cultivation and trading of watermelons in different cities of the country via seasonal migration. There are situations in which Roma boys from nearby communities are hired by Rudari to do work on a daily or monthly basis. Rudari girls are married at age of 13 to 14, and they are under strong pressure to perform domestic chores for the groom’s family. This is associated with restrictive mobility and controlled social relations, school dropout and functional illiteracy.

– P., a semi-traditional Roma community who identify themselves as “Gypsies”⁶ living in close proximity to a national highway linking two cities from Transylvania. There is relative mobility of community members and consequently, greater potential income. The main activities of children include occasional work for wealthier Romanian and Hungarian families, extracting and selling sand from a river, and berry gathering.

– Ba., a rural community of 1,200 members from Transylvania, representing one third of a relatively large and isolated village, with few and expensive transportation means to any city. The community has a recent history, linked with the agricultural potential of the village during communism, which attracted Roma from surrounding areas. Without any employment opportunities following the collapse of centralised agriculture and with a 30 per cent illiteracy rate, the Roma are currently highly dependent on social security benefits. According to the interviewed Roma, the humanitarian assistance from religious leaders and limited social programs have divided the community. The school attendance rate is low even though the school is close by. The few persons who have graduated from vocational/technical schools in the city are now back home, unemployed.

– R. neighbourhood, an urban, extremely poor, mixed community from a city in Transylvania. Differences between poor and excluded Romanians, Hungarians and Roma tend to be small. School attendance is relatively high, but with few children attending vocational or high schools afterwards. There is a high teenage pregnancy rate and a “phenomenon” of children’s idleness (children who are neither in school nor working). There is no homogeneity of occupations that children here might resort to—they vary from domestic work, seasonal work in farms outside the city or collecting recyclable materials (metal and paper) to replacing parents in the community work they have to perform so as to get social benefits. Professional possibilities for both boys and girls are limited.

– Interviews with children who sell newspapers, aromatic plants or woodcrafts in the streets of a developed city in Transylvania. They are older girls from the orphanage, now living independently (in a deplorable night centre); or they come from neighbouring villages to sell plants (a job they consider to be on the border of begging) or migrant boys from highly isolated villages.

Roma are far from being a homogenous ethnic group. It is difficult to represent their diversity in terms of region, level of traditionalism, occupation, levels of education, or views on the value of work. There are, nevertheless, major dilemmas about “what makes a (Roma) community”. The dynamics in Roma settlements (defined geographically) may render problematic the tendency to ascertain the relation between community and children’s work.

Children from the same community tend to engage in similar activities (including work). Interviewed parents speak about a collectively shared view regarding the role of children in the daily struggles of families and the degree to which they respond or not to expectations. A Roma community often feels the need to distance itself from another in its close proximity.⁷ In this demarcation, the status of children and the choices with regard to education are often social markers or symbolic boundaries in Cohen’s (1985) terms. One Roma community may criticise another one for not sending their children to school and for keeping them in deplorable situations that may entail stigmatising work (e.g. begging). This relational process corresponds to what Barth analysed as “dichotomisation” (1969). These concealed borders inside Roma communities may bring with them the risk of oversimplification when speaking about “Roma communities”.

The majority of the “communities” where the interviews took place are from Transylvania, which is the region with the biggest population of Roma (NIS 2002). The most traditional ones (in terms of occupation, language and customs, such as the Gabori, Kalderari and Cortorari) have not been included in this study. I thought that in many respects these groups have been more successful in “finding their way” in post-communist times. Their practices may justify a study in its own right. There appears to be anthropological interest for traditional Roma even if not focused on children (Voiculescu 2002; Troc 2002; Troc, in Fleck and Rughinis 2008). My knowledge regarding these most traditional social groups of Roma is limited. However, they represent a small minority now in Romania.

Roma working children experience many forms of discrimination and vulnerability inside and outside their communities. The research process aimed to empower Roma working children, who have a marginal position due to a combination of factors, including ethnicity, age, gender, poverty and at times disability. The main idea underlying the fieldwork was that it is misleading and unfair to understand Roma children as powerless victims of actual situations that they are living in. Apart from the structural constraints that shape their lives one should acknowledge children’s own “agency”, their capacity to make sense of the world they inhabit and their ability to talk about how they understand both work and education.

The research was undertaken with a concern for ethical issues. Often the ethical aspects involved in research with children tend to be reduced to informed consent. While this is a relevant and sensitive matter indeed, there are other ethical issues to be considered. This study tried to incorporate ethical aspects in the whole process. First, the selection of the topic aims to bring one of the social groups that is usually overlooked closer to academic and social interest. Second, the data collection and interpretation was undertaken in ways that respect the children’s self-expression and values their perspectives as epistemologically relevant. The interviews took place in settings children were familiar with. The research deliberately avoided the use of questionnaires and other written data collection instruments.

One of the dilemmas that I tried to balance was the unequal distribution of power that is given by age, education, ethnicity and class. For handling the age differences without reproducing the pattern of subordination toward adults I adopted the strategy proposed by William Corsaro (2005). According to this method, while not declining its belonging to a different category the researcher needs to meet the role of the adult who is different and sincerely interested in the children’s experiences (Haudrup 2004; Corsaro 2005). The interviews in a semi-traditional rural community were undertaken by a Roma person.

5. Roma working children

5.1 The legacy of the past

The roots of children’s problems are not always to be found in adult problems and vice versa. Seeing children as separate from the communities and societies they live in is a great injustice. Children have always contributed to the welfare of their communities and still do. Their actions are not separated from those of the larger neighbourhoods, communities and society they were part of. Despite the attempts to seclude childhood into an apoliticised private sphere (Mayall 2001), children’s lives (and Roma were far from an exception) were touched by political and public transformation.

During communism, egalitarian policies promoted free education, housing and employment, for Roma too. In the cities Roma lived in mixed communities and there were no major discrepancies in living standards in comparison with the majority population where persons had several children. Roma from the countryside did work in agriculture or in cooperative handicraft (some traditional occupations were gradually restricted in later communist years).

Recent statistical data shows that 10 per cent of Roma still undertake traditional occupations, whereas 37 per cent practice qualifications acquired through the formal educational system (Chiriac and Constantinescu 2007: 13–15). Under these circumstances, prevailing arguments that refer to Roma children’s engagement in traditional occupations that are transmitted generationally (Cace 2002: 28) need to be revised. This is an issue to be discussed in the next section.

Almost half of the Roma live in the countryside; statistics show that only 23.8 per cent own land and 41.4 per cent have a garden (CASPIS 2002). In these circumstances the only sources of income are sporadic and poorly paid day labour and, not rarely, stealing (CASPIS 2002). It is under these circumstances that, unlike children from majority population, Roma children working in agriculture tend to work for other landowners than their immediate family. This raises questions about the social risks the children are exposed to.

Many of the families who lived in the cities during communism needed to move out given the increasing unemployment, high living costs and, often, urbanisation pressures. They frequently settled into marginalised and often overcrowded neighbourhoods. The internal dynamic of previously established communities transformed, following the gradual or rushed arrival of other families. This is a recurrent theme in the discussions with Roma from the R. neighbourhood, who recall families that were “relocated” in social settlements. Conversely, Roma from the rural community of Ba. speak about several families who joined their relatives in the countryside after living for years in the city.

New forms of power, influence and (dis)trust emerged. Children involved in “worst forms of child labour” (children borrowed for work, trafficked, begging) tend to come from the newly formed communities. Previous social-capital explanations of child labour link it with a low level of social inclusion of children in their communities. William Myers (2001) argues that children who are improperly connected to their communities are more likely to enter forms of labour that may be qualified as abusive. In analysing the consequences the process of family replacement a connection with an apparently distant example supplied by M. Liebel (2004) might have some explicative value.

When referring to the sudden increase of child labour in Africa, Liebel (2004) finds that the causes are linked to the colonial experience. Whereas any connection with Roma from Romania may risk being charged as hazardous speculation, one may still use his analysis in order to generate a hypothesis on Roma children’s work. Both African and Roma children have been active members of their communities. Liebel (2004) argues that when the stability of a community is high and it operates on self-regulated principles children are gradually introduced in work by assuming tasks that are considered appropriate to their age and development.

Maintaining the necessary proportions, what links colonialism in Africa with the new social transformations related to Roma in Romania is an unexpected manifestation of power. Communities were destabilised by new forms of poverty and structural changes (crises in housing, land, unemployment). Roma communities are not self-sufficient: with poor resources, Roma need to turn to the majority population or the state. Such relations are ultimately power relations. “Exploitation takes place in all societies in which there are unequal power relations, and the powerful section of society is able to tap the productive potential of those excluded from power”(Liebel 2004: 202).

All these transformations are collectively experienced as disruptive. Ultimately, Liebel argues, the most harmful forms of work children undertake are not a result of cultural backwardness but are matters of poverty and power (Liebel 2004: 100–1). The same arguments were articulated in the discourses on child sexual abuse and the feminist discourses that argue that the status of women and children are intertwined given the power unbalance society allocates to their status. In a more radical discourse childhood was, ultimately described as a state of oppression (Kitzinger, in James and Prout 2006: 182). The situation of Roma working children thus speaks about transformations both in Romanian society as a whole and in Roma communities.

5.2 Work, its benefits, risks and children’s views

On the whole, neither Roma, nor Romanian child labour resembles the industrial forms more often associated with Victorian England or with today’s textile industry in Southern Asia.⁸ In the main it refers to children who combine (limited) school attendance with working on small family farms. According to official data, almost 90 per cent of all child labourers from Romania work in agriculture (NIS 2003). For the majority population it is not school attendance but school performance that is affected by work. School dropout makes an appearance at a later stage, causing low educational achievement due to work undertaken and/or another situation (poverty, family crises, no school nearby, etc.).

For the Roma population the situation is more critical. Whereas majority-population children are more likely to work their own land together with the family,⁹ Roma households are poorer so children will need to seek work further afield. Given this fact the risks of exploitation and abuse are implicitly higher. Besides seasonal work in agriculture Roma children (especially boys) work in construction. National legislation forbids work in construction below the age of 16 and this adds to the already existing risks. Working at heights, exposed to toxic substances and the multiple disadvantages of being Roma, young, and “the least qualified” member of the team—all this makes them a particularly vulnerable social group.

The collection of recyclable materials (paper and, especially, metal) from public spaces but also from soil, water and garbage exposes children to a social stigma and also to health risks (Fassa et al. 2000). At home many children are engaged in childcare and/or domestic work. Depending on age, boys might also do such work.

Evidence related to the health consequences of child labour is poor (Scanlon 2002; Hesketh 2006). There are few large-scale longitudinal studies and there are many methodological problems to be overcome. The direction of causality is difficult to establish. One cannot estimate for instance, to what extent work damages the health of child labourers or whether the fact is that children who enter the world of work tend to have poor health anyway (OECD 2003).

Roma working children speak about the physical and psychological harm caused by excessive work and improper conditions. Physically they may be injured by exposure to dangerous substances (for example, lime burns in construction or by inhaling dust and toxins at a chicken farm), working at heights, in an accident with a car (for children who wash cars or sell newspapers at crossroads), by heavy lifting, sunstroke (for children working in agriculture) plus unhealthy sleeping and living conditions (for those who sell water melons in cities, close to a main road and in the open, without security measures and being exposed to bad weather). Low nutrition, poor hygiene and irregular sleeping hours when at work are also among their more tacit complaints. If attending school, teachers refer to working children’s fatigue and their low ability to concentrate.

From a psychological point of view, the interviewed working children tend to have ambivalent feelings about themselves. On the one hand, some are proud of being able to do more than majority-population children do (especially if they work in construction, in agriculture or farming, which involves heavy workloads). Parents may also accept a more emancipated form of behaviour from children who contribute to the family’s income. On the other hand, interviewed children seem also to internalise a depreciatory feeling of being different from their peers who are not working. There is wide agreement in recent studies that despite the poor longitudinal research, girls’ domestic work is less socially visible and may lead to emotional burnout and a sense of alienation.

Interviewed children working for an “employer” appear to be more exposed to ill-treatment and neglect. “Employers” do not act in loco parentis and children internalise what the employer considers to be typically adult behaviour (smoking, drinking, even visits to prostitutes): children working in adult-dominated settings join the smoking breaks and part of the payment is scattered in small amounts as “cigarette money” during the day. In many cases child labour could open the door to a premature adulthood and anti-social behaviour (e.g. children who collect metal may steal iron or even enter into a metal-stealing network).

The reasons why a person may offer a job to a child are manifold. In agriculture children over 13–14 years old can obtain the same amount of money as their parents and during the agricultural season parents and children need to make use of this opportunity. Sometimes the “employer” may believe he or she is helping the child and its family (the employer is a relative, a neighbour or an old family acquaintance). There are also situations when the intention to take advantage of the child driven by economic profit-making seems explicit:

“So I brought this older one here at the weighing machine, you know?”

“Who is the child?

“He’s the child of one boy from my village. He gave it to me […] He’s helping a bit… […] cause it’s very hard for them, they’re poor…[…] two years ago I’ve had his brother along, but he grow up and all he wants is going in the city […].”

(Man, Rudar, 35, watermelon seller)

Children who work in a seemingly formal way (as newspaper sellers being “employed” by a representatives of a printing house), will display a high level of confidence in their employers—despite the fact that some conflicts may also occur:

“I sell newspapers, even if I am only 12… I told my boss that I’m 16 because I look like I’m 16—and he believed me.”

(Maria, 12, newspaper seller for a year)

“He (the employer) promised to help me get a high school diploma. He knows somebody.”

(Catalin, 17, newspaper seller for five years)

Not all children who are not attending school are working. There are also many “idle” children (who are neither at school nor in work). Their families/ neighbourhoods are extremely poor and have no work opportunities.¹⁰ In the worst scenario “idle” children run a high risk of entering the worst forms of child labour (e.g. criminal activities, trafficking, prostitution). In extreme situations some Roma families also practice child-trafficking themselves.¹¹ As regards external trafficking, different patterns exist:¹²

–– children migrate and work as informal and irregular migrants together with their families;

–– children live abroad with a close relative for financial gain by involvement in small criminal activities;

–– there are unaccompanied children left abroad by their parents or other adults who did not declare that they had children when leaving the country.

What do children think about their working lives? Talking to children about their work does not always give the uniformly depressing feeling one might expect. Children do not see work as being entirely positive or as entirely negative (the same conclusion was arrived at by A. Invernizzi 2006 with reference to Portugal). Most do see benefits in the work they do (like seeing the city for those living in the countryside, earning money to buy “what they want”, having more independence).

If compared with non-Roma,¹³ Roma children are more likely to refer to the negative side of their work. Roma children living in the R. urban neighbourhood are more likely to say that they enjoy opportunities to work as a means of earning money. They are not proud of working but do “not feel ashamed” – which gives reference to their learned feeling of being disadvantaged. It is important to note that the “Western” practice of part-time jobs for youngsters is really at its very beginning in Romania. There is implicit understanding that poverty is the only reason children under 16 may seek work.

However, children try to negotiate the reasons of entry into work in terms of individual autonomy. They tend to internalise the decision of entry into work as their own. It is only at a later stage that a larger family dynamic becomes manifest. For instance, Ştefan, a Roma boy of 13, working at a poultry farm in an adult dominated setting, highly stigmatised, speaks about his work as follows:

Ştefan: If I like work, I need to go. It’s not long until I’ll get my ID card (at 14, nn). I have to build up my own room. I have to work myself.

Q: This is what you think or what you were told?

Ştefan: This is how my mother told me, my grandpa, my uncle who’s living with us: to think about going to work myself, if I want to. He’s not obliging me, only if I want to… to go and make some money myself, to by some stuff for myself ”.

(Ştefan, 13 years old)

The marginal position of the child in the family (in grandparent’s custody, as in the case above; just married, living with the family of the husband), brings along a set of limitations that lowers the child’s autonomy. In these circumstances, work becomes simultaneously a form of power and control exercised from above (by parents or other care takers) but also a form of gaining voice and negotiation influence by children themselves. In such circumstances, children internalise an ambivalent sense about their work as being both autonomous and stigmatising. It is under these circumstances that the fight against child labour may cause more harm than good to the children themselves, if other enabling circumstances for children to negotiate their position in the family are not available.

Interviewed children who beg also try to define their work in terms that are more socially acceptable. Selling items for which there is no genuine demand is a strategy that reduces the social stigma and meets its economic purpose. Among the interviewed parents and the children themselves there seems to be consensus that smaller children are more successful at begging or selling items on the streets. Given the reduced market for various products that may be sold, the age of children who are involved is now decreasing. In maximising their potential (Romanian and Roma) families engage younger and younger children in street selling. Interviewed young people who are less successful in selling recall the older age of the former peers at the time they started to sell on the streets compared with the age of the children who are selling at present.

By and large, the children engaged in street trading included in this research can be described as belonging to three groups. First, there seems to be a group of children who used to beg when younger, but who develop new strategies for reaching the public, such as selling various and sundry items like sweet basil and small religious icons. These are things that people do not usually buy, but they are used as pretexts to approach potential benefactors. The messages children articulate are about need—not about the items to be sold (for which they will not request a specific price). Depending on their age, children in the same family may beg or sell in different parts of a city.

Second, there appears to be a group of children, usually boys, who began and later continue to work as street vendors. These children are usually Romanians from isolated mountain areas such as Apuseni who came to the cities at the time when there appeared to be some demand for their products, such as small wood items (spoons, hangers, etc.). In their specific case, there is a legal document that legitimises their work as a recognised program of rural development. These children tend to live in cities in groups, and go home occasionally for seasonal agricultural work. Many remained sellers longer than initially expected, for as long as six to eight years. During this time, their economic role on the market often become diminished, to the extend they may be faced with a dilemma regarding their presence on the streets: “Some say we are beggars. I don’t know what to say…” (Mircea, 18 years old). Periods of religious celebrations are described as more successful as children often receive money as charity.

Third, there is a more independent group of children who sell newspapers in established areas of larger cities, especially in the proximity of important intersections, where traffic lights generally provide the rhythm for their work. Children may work at the same intersection, with individual delimitations of clientele dictated by traffic lanes. Roma and Romanian children may share the work at these locales.

In the first two situations, both parents and children are aware of the economic benefits that results from children being of younger age and use it as a social tool. To a certain extent this can be described as the instrumentalisation of childhood. The last group seems more autonomous: the majority of interviewed children selling newspapers are living independently at night shelters and have their own economic interests, such as saving money in individual bank accounts. A child’s success may improve with age as well, especially if they form close relationships with clients in the same place over several years.

One should not be deceived by a child’s tendency to see their working life in positive terms, however.¹⁴ Living in more or less difficult situations, children— like anybody else—have a psychological need to develop internal strategies with which to rationalise their often deprived conditions. In these circumstances, it becomes understandable why many working children, in spite of the objectively difficult work they perform, still find their work “not hard at all”, or even satisfactory, and voluntary, too: “I work only when I want to and how much I want to” (Ana, 13, a newspaper seller).

Amarthya Sen (1999) looks at the same argument as one of the three limitations in utilitarian ethical theory, namely “adaptation and mental conditioning”:

. . . deprived people tend to come to terms with their deprivation because of the sheer necessity of survival; and they may, as a result, lack the courage to demand any radical change, and may even adjust their desires and expectations to what they unambitiously see as being feasible.

(Sen 1999: 63)

Consequently, one could argue that—in Sen’s terms—working children, like any deprived social grouping, can adjust psychologically to persistent deprivation. Given this, the apparently positive attitudes toward work and the destitute condition of Roma working children should not be understood as a lack of sensitivity, or indeed, an increased resistance, to deprivation.

3. Risk factors inherent in entering child labour

There is a large body of literature on the economic and cultural factors that cause and support child labour. However, despite many economic analyses on the topic, it is often difficult to find clear causal relationships between economic conditions and the presence or absence of child labour. Below is a somewhat artificial attempt to identify and organise the risk factors of child labour—they are often interrelated and difficult to categorise. Some of the risk factors, such as parental education, are more external to the individual; others less so, such as the power dynamics of the family, which are clearly visible at the household level. Nevertheless, merely classifying causal risk factors does not assume that child labour can be easily solved, as there are often deeper structural constraints behind the problem.

When it comes to child labour, one of the most frequently discussed issues is poverty, though one could argue that poverty is only one factor that perpetuates those discriminatory practices that lead to a child’s early entry into work (Weiner 1991; Murshed 2001). Ethnicity is another cross-cutting theme among the casual variables summarised below. Being Roma does matter when talking about poverty, education, community and, often, about family.

3.1 Factors related to poverty

According to Roma children and their families, the main reason why they work is to contribute to family income. Due to extreme poverty and high unemployment rates, many Roma families in Romania live on child support and social-security benefits. Even if education is officially free, Roma children and parents speak about the cost of clothes, food and transportation costs that they cannot afford. The fact that children, especially boys, speak about keeping part of the money they earn for their own living expenses is also an expression of the family’s inability to provide for the children’s basic needs.

However, research demonstrates that the relation between poverty and child labour is often not linear (Bhalotra and Heady 2003, cf. OECD 2003; Ghinăraru 2004). There are many different types of poverty (e.g. household poverty, community poverty, regional poverty) and each poverty type on its own may or may not predict the prevalence of child labour. Whereas children tend to work in poorer households, small-holdings may actually increase the probability that children will work (Ghinăraru 2004). In the present study this seems to be the situation in the Rudari communities, where land ownership is common among most residents.

As stated earlier, Roma are much poorer than the majority population in terms of income. However, a much wider definition of poverty, understood as “capability deprivation” (Sen 1999), is better able to link the economic conditions of Roma with their social status. There are other factors, not just income poverty, which keep Roma families from obtaining the same social standing as the majority population. When possessing the same level of income, a Roma child is more likely to work, whereas a non-Roma child will more likely go to school. This fact says a lot about the broader context of deprivation that the Roma population continually face.

Although an increasing number of families from the majority population of Romania also experience problems of unemployment and poverty, solutions for them are often very different. With the majority population, one or both parents may emigrate to find work, whereas with the most deprived Roma families, parents are more likely to remain at home with their children. The same pattern was seen in relation to the Roma in Albania (De Soto et al. 2005). The main reason why Roma from Albania do not emigrate (which I found for highly deprived Roma from non-traditional communities as well) is that they lack both the initial money and social capital (De Soto et al. 2005). If they do, they tend to emigrate together with their entire family, including children, and have very little trust in the institutions aiming to “mediate their positions”.¹⁵

3.2 Factors related to education

Previous research shows that the relationship between school and work is not uni-directional, therefore it is not clear if “child labour discourages school attendance or if it only lowers the quality of school attainment” (Sedlacek et al. 2005: 2). Moreover, it is nearly impossible to determine whether children remain absent from school because they are working, or whether they are working because they do not go to school.

Globally speaking, for one category of children it is certain that the work they do prevents them from attending school. For another category, however, it is the education system that pushes them out—and only then do they start working. However, perhaps the largest category of children is made up of those who combine school with work. This, according to Lieten and White (2001), is one of the aspects usually overlooked by public-policy programs. Studies of child labour tend to assume that children who work do not go to school, whereas studies of education often ignore the very fact that many children work.

This issue is particularly relevant for designing public policies, as interventions aimed at “increasing enrolment are different from those aimed at raising the productivity of time spent in school” (Sedlacek et al. 2005: 2). The next section will try to bring together some of the barriers Roma children face in their education in Romania.

3.2.1. Enrolment

In general, the number and quality of schools and kindergartens in Romania is low. The data relating to schools in general do not distinguish between predominantly Roma institutions and other schools. In spite of the fact that new schools are being built each year, many children continue to study in deplorable conditions. Due to demographic changes in rural areas, many villages do not have enough children for schools to function. Under these circumstances, there is an interest in enrolling Roma children just to prevent schools from being closed. This, according to Michael Stewart, is one of the reasons behind the high enrolment rate in Romani’s rural kindergardens (55% in 2008) (Stewart, in Fleck and Rughiniş 2008: 10).

As a direct result of the decentralisation trend in the Romanian educational system, which gives municipalities and local communities more responsibility for funding part of education as well as alleviating general poverty, one can assume that the more Roma children are present in a school the lower its material standard. The few exceptions are a small number of new schools being constructed in Roma communities with non-governmental resources, such as those funded by PHARE projects or other international organisations.

3.2.2 The quality of education

The quality of teaching staff is less satisfactory in more isolated villages. Qualified teachers are reluctant to go there because of limited transportation and because the salaries do not weigh up to the working conditions. Nevertheless, there are many Roma teachers who are qualified to teach the Romani language and work as school mediators. Although many training sessions and resources have been devoted to the education of Roma teaching staff, drop-out rates among Roma children continue to increase.

The Roma parents that were interviewed perfectly understand the constraints teachers face when they try to comply with different pressures placed on them by parents from the majority population. Most are aware of the limitations of working with increasingly defiant children, sometimes in overcrowded classes. Yet they also have positive memories of their own school days. There are still situations when Roma parents tend to have expectations that sometimes exceed the professional duties of a teacher:

Teachers should make explanations until the child understands. When I was a child, teachers didn’t let you go home if you didn’t understand something! They kept you on after hours and explained everything! Now, they’re just waiting for the break.

(Mother, aged 32).

Even if there is a governmental recommendation concerning the elimination of school segregation, there are still many schools in Romania that have a Roma majority. One of the problems that comes with segregation is the low quality of education, the poor motivation of children and teachers alike, and low educational aspirations. Moreover, simple attendance is not to be confused with good school performance (White and Lieten 2001). Roma children may find themselves almost illiterate at the end of four primary years, which increases the likelihood that they will eventually drop out of school.

However, even when schools are of good quality, and are close to home, this does not necessarily mean that Roma children will receive a good education. Lacking the time and physical space to prepare for lessons at home, working children are often tired at school and less able to concentrate on their studies. Low nutrition, inadequate housing and poor family support for school activities are also barriers to good-quality education. Once again, it is difficult to establish a direct causal relation between participation in work and the presence or absence of education barriers.

At this juncture there is not enough evidence to suggest that children are poorly fed and tired because they are working, or, conversely, that they are working because they are very poorly fed. Interviews in the rural community of Ba. made it clear that for some families, income or material goods earned by children are vital to their existence. In those instances, preventing children to work without ensuring other means for family survival may cause more harm than good.

3.2.3. Attitudes toward education

A large number of Roma parents have had eight or ten years of education, attained mainly during the period of communism. Even if many are functionally illiterate today, during communism they were included in different levels of the educational system. Later on, many of them received workplace certification, and a decent, long-term, salary was guaranteed. Parents know their professional achievements were largely thanks to this former system, and they also know that such a protective system does not exist for their own children. During the transition, many developed a strong sense of helplessness that is transferred form generation to generation:

It was much better when I was of her age—there were four salaries in the house. […] Now, people don’t have jobs, so they are living only on social security benefits. In winter there is nothing here you can do but wait about.

(Woman, 45 years old)

Both the parents and children that were interviewed say that education matters. They have a strong sense of what is socially desirable, and invariably use middle-class discourse when referring to the value of education. To a certain extent this attitude is typical of many other disadvantaged groups. They tend to see education as a vague and long-term project that cannot be divided into having short-term goals (Olthoff 2006). Roma parents, and their children too, are inclined to have this attitude. However, at a latter stage they may establish a contrary position—which tells us more about the dilemmas experienced by many children, Roma and non-Roma alike.

According to C. Ghinăraru (2004), many disadvantaged people stop seeing school as important for them as they are less exposed to the principles of the new labour market and secluded from a world where education matters. Thus, the high number of working children in isolated Romanian villages reflects a disparity in educational achievement due party to the lower educational aspirations that result from rural isolation (Ghinăraru 2004).

On the other hand, the benefits to be gained from education—especially secondary education—in today’s Romania are relatively meagre. Many Roma opinions are grounded in such logic. Those parents that were interviewed know that education is not always the answer to poverty and unemployment and this may add to the reasons why adults do not invest time, energy or money in schooling. They have a deep sense that society does not reward merit as it should and that a confusion of values and economic polarisation has made education problematic for a large proportion of Roma and majority populations.

In addition, success stories coming via education are extremely rare in Roma communities. Young, educated Roma are often unemployed, and struggle with the same difficulties as the uneducated. Not without reason, the repeated saying of the unemployed university graduate is often invoked to support the argument:

What’s the point of going to university when you see that people who have graduated are still unemployed, working in construction or cleaning floors!?

(Ana, 12 years old)

Another reason why school is not regarded as a profitable investment in the long run is the increasingly higher educational qualifications needed to enter a restricted labour market. “Credential inflation” is a term that best describes this situation. This theory was developed by Randall Collins in the early 1960s, and refers to the current devaluation of school diplomas. Similarly, Boudon described an “inflation spiral” (Boudon, cf. Elster 1992: 48). For many Roma, the low returns from investment in secondary education undermine any initiative to continue education. This attitude tends to be characteristic of a large part of the general population, especially in rural areas.

Attitudes toward girls’ education in those communities included in this study were also found to be ambivalent. In the case of non-working or above-average urban Roma families there is an incentive to educate girls so that they obtain reasonable and “light” work afterwards. Poorer parents are more aware that such opportunities are less obtainable. Roma see labour-market discrimination against women and think that working options for girls will be limited to being a portress or perhaps an employee at a poorly paying clothing plant. On the whole, those parents that were interviewed did not seem influenced by patriarchal and oppressive attitudes towards girls in the sense that one might expect. Their attributes are, rather, internalisations of a discriminatory, and often patriarchal, labour market.

Depending on their age, working girls from the urban neighbourhood of R. who are still in school are more assertive and generally have the goal of working in similar-level jobs, for example as hair stylists or shop assistants. One may argue that their career prospects are more “realistic” since they know their economic means and they have numerous role models who have succeeded in their educational efforts.

Many policies that have attempted to raise the level of school attendance of Roma have focused on needs to improve the curricula in order to incorporate children’s different cultural experiences. Where it is beneficial to have more inclusive courses, there is still some scepticism about their ultimate value. According to the same theory of credential inflation mentioned above, parents generally know that education is not an end in itself. They send their children to school not primarily for intellectual reasons, but so that they may get a decent job in the future; or, in other words, “the reasons for going to school are extraneous to whatever goes on in the classroom” (Collins 1979: 192).

In the final analysis, seeing “the problem of child labour” simply as a matter of negotiation between schools, families and children is not adequate to provide an ultimate solution to the problem. By including the labour market in the equation and revisiting the function of schooling in relation to employment opportunities, may be a more constructive approach. It may sound like an overly large enterprise, but one must acknowledge that schools do not function in a vacuum. Unless energies are directed towards the labour market, schools will continue to make false promises to the children of both Roma and majority population alike.

3.3. Factors related to values and norms

The typical way in which many Romani children enter the world of work is by first learning the relevant skills from their parents. There are still many examples in traditional communities where children directly take up the occupation of their parents. This was the case with Roma families who were self-employed during communism and later became successful in finding an economic niche in a new economy. Many of them face fewer problems when adjusting to post-communism, and, according to Fleck and Rughiniş (2008: 56–8), many do not identify with modern Roma groups and are actually proud of being “gypsies”.

The assumption is that because of higher levels of poverty and exclusion, non-traditional Roma who used to be employed during communism and who now rely on social security schemes, are more likely to have children entering the workplace in ways that may be detrimental to their social development. I argue that non-traditional Roma children are more likely to do unqualified jobs as part of a family strategy to resist poverty, not because there is an occupational tradition to be maintained or a profitable family business that they can contribute to.

The most prosperous group this research included seemed to be the Rudari. From fieldwork interviews, it appears that the work of Rudari children appears to be deeply incorporated into their larger economic livelihoods. For example, children participate in the preparation of the watermelon seedlings in early spring, and later, cultivating and selling of them in the summer. For part of the year, many Rudari adults are seasonal workers in Spain. During the summer, for two months at least, whole families work as street vendors in the country’s cities. Many dwell in small towns, in compact neighbourhoods, with social boundaries that restrict interaction with other Romanian and Roma populations, especially when it comes to dating or marriage.

However, several interviewed Rudari have chosen to abandon some Rudari values like early marriage and buying the bride, and indeed are now critical of them. Others qualify the practices as deplorable, though continue to adhere to them. The consequences of doing otherwise could mean assuming a dissident position in the Rudari community and a marginal one in Romanian community.

Girls are especially vulnerable in these situations. Children’s work goes along with early marriage. Often girls are married at the age of 13. Living with the groom’s family, a girl’s status may resemble that of a young domestic servant, for they live in a strict and controlled environment. They internalise traditional gender roles at an early age:

Once you get married, you don’t go to school anymore! You are now married, so you stay at home to do the housework! If your mother-in-law sees that you aren’t working, she’ll drive you out! […] Some mothers-in-law are really awful. […] You have to be standing and on the go all day long. If she catches you sitting down for a breather, she’ll go crazy! She has demands on you, even if you are young […] She wants you to do more than she’s doing. She wants you to take over her hardships […] To get up at five in the morning, feed the birds, to cook in silence until she gets up at 10.

(Cristina, 13 and half years old, married for seven months)

As in other traditional societies (Doftori 2004), a strong focus on strict gender roles and values of seniority give certain persons authority over children. It is in this context that children’s work for the family welfare becomes a social expectation. The work power of Rudari girls thus appears to be transferred from their original family to the groom’s. Limited social interaction with parents or previous friends (unless they are married) means that a strong sense of isolation and sometimes depression may accompany the first years of marriage. Things do tend to change when the new family has its own children and gains more autonomy, even if it remains in a multi-generational household.

The interviewed Rudari adults argued that the presence of an extended family is supportive and is something other ethnic groups do not have. They know that it is hard for a young couple to live on their own, especially when they have got a poor education and live in a rural area. However, whereas the family does support a young couple as regards to building a home and cultivating land, it also pushes them into marriage in ways that are not always explicit: “The kids loved each other. What shall we do?” Back home, when an older girl gets married, her younger sister or brother takes over a part of the housework, which may often include taking care of other younger children and also working on the land.

In many non-traditional Roma communities, as well as in traditional Romanian communities, child labour appears to be culturally acceptable, except in its most extreme forms. Internationally, it is recognised that child labour would not be so common if it did not benefit from social and cultural approval.¹⁶ There is a general assumption in traditional Romanian communities that work is good. In this culture, children are expected to contribute to the family’s earnings. Many Roma and majority-population families share the belief that early work shapes character and makes children better prepared for life. What is problematic, though, is that the boundaries between assuming small responsibilities (which is helpful to development) and child labour (which is detrimental) is often vague and not well understood.

Apart from most traditional Roma families, child labour will tend to be an issue arising from a deprived economic and social conditions than by deeply rooted cultural norms. An informal labour market based on semi-skilled manual labour reinforces earlier social expectations regarding the role of children in the workplace. In these circumstances, late childhood is socially constructed as the age of maturity. At the age of 14, children receive identity cards; this is also the age when they finish elementary school. For rural children, the continuation of education would mean higher transportation and housing costs. Under these circumstances, the age of 14 is the doorway to adulthood—and entry into the world of labour.

In addition, the mobility of some Roma communities as the result of external or internal migrations to do seasonal work influences the early entry of children into work. Children generally leave with their seasonal worker parents as there is no one left at home to provide care for them. When Rudari mothers leave for agricultural work in Spain to pick olives or oranges, they may bring along their adolescent daughters. This allows for greater family control of the children, but also maximises their economic gain as young girls often work alongside their mothers.

The practice of leaving school approximately two weeks earlier than others for different labour tasks is also found among Roma from rural areas. Even if parents may not see all this as an important loss, the educational system is competitive and such practices do still put Roma children in a position of disadvantage when compared with the rest of the students. Besides, both Roma and poor Romanian parents have a tendency to withdraw their support from children who are less than successful at school (Stativa 2004).

He didn’t like school. What could I have done? At the beginning I beat him, but after a while I realised that there’s no point […] that maybe I cause him something and it will be my problem all over again. […] Now he’s 15. He can read individual letters, yet it takes him hours to read a page. But when counting money, well, he’s an expert!

(Father, aged 33, Rudar)

Children are not passive when it comes to working or making work-related decisions. It would be a great injustice to connect child labour directly with parental culpability. Children’s entry into labour is habitually mediated by the way children themselves understand poverty. From an early age, many Roma children and poorer Romanian children have experienced deprivation, and have sought out solutions to be able to resist and transform it. The tendency to victimise working children is an implicit denial of children’s own agency in the practice.

There are not only “Roma values and norms” at play when it comes to child labour. In the final analysis, there would not be so many Roma working children if there was no demand for it and a major degree of cultural “acceptability” coming from those who benefit from it. Labour demand may take many forms. It may include subsistence agriculture based on family farms (Ghinăraru 2004) where the use of undeveloped technologies creates an informal market for child labour. The ageing population of the countryside also increases the need for a seasonal workforce. Roma often do such work, and, as stated earlier, their children will also.

It is noteworthy that many working Roma children from rural areas have some experience of working—with their families or alone—for wealthier Romanian or Hungarian families. So they may have internalised an ethnicity-based, subordinate social position at an early age. Roma child labour is still accepted to a major extent by the general Romanian population. The use of Roma children’s capacity for work, even if in isolated cases, appears to have a high degree of social acceptability for children, their families, employers, and also for the authorities, who may well turn a blind eye.¹⁷

Final remarks

Research on children’s work can hardly be neutral. For a long time, the concept of children’s work carried a political, moral and ideological burden. As stated in the first part, there are different positions one might take with reference to children’s work, including the extremely protectionist discourses that consider work “a pathology of childhood” (White 1999), to a more liberal approach, which sees work as a right that cannot be denied to children (Liebel 2004).

In the final analysis, what are the implications of one or another position for Roma working children? This article was initially intended as a policy paper. However, this last part does not propose policy recommendations, but, rather, explores the dilemma outlined above.

On the one hand, it is difficult to state that Roma children should not work, as long as for some, work provides the means for a living and also for school attendance. Placing school as the legitimate place for children to be is equally problematic: it becomes a normative and ultimately, a political statement with regard to a minority. Still, there is a long way from here toward advocating for Roma children’s right to work, also.

On the other hand, in many developed countries, most of the children who work do so while being motivated by a “consumerist” need or to gain a sense of responsibility and to enrich their experiences (Morrow 1994, 1996; Liebel 2004). More recently, as stated before, children from developing countries organised themselves to demand the right to work in conditions that are decent and nonexploitative (Liebel 2004).

Roma children appear to be situated in between the two worlds. They live in a competitive world where education matters and where, from an early age, parents strive to increase the academic chances of their children. In such a world, the part-time work of children may be part of a middle-class educational strategy.

However, in such an environment Roma children, come into view as a non-Western model that tends to be rather described than interpreted, out of the precaution of not forcing non-western notions of childhood into Western moulds (Weiss 1993, cf. Liebel 2004). In this respect, Roma children may be described as inhabiting a different model of childhood, a different cultural world where work is legitimised as duty, professional training or gender socialisation.

Despite the risks of qualifying the discussion on education as normative and inappropriate for a qualitative research, and despite the arguments coming from South American children’s movements for decent work, this research points in a different direction. Given the highly competitive world of education that children are now involved in in Europe, this article argues that advocating Roma children’s right to work, be it in the name of children’s agency or of an essentialised Roma culture, may risk carrying the subtext of racial and class prejudice.

Many Roma children work in a context of deep deprivation and with a scarcity of alternatives. Like many children in developing countries (see White and Lieten 2001), they cannot choose the nature or the duration of their work (full time, part time, evening hours, etc.). For many of them, work is often an essential part of their lives, not a transitory or contextual practice to be used for a specific purpose. Under these circumstances, the debate around the value of work for future life achievements is weakening. There are only very a few Roma children working to learn a trade. Their large majority is undertaking various works that are occasional, poorly skilled, and important in their families economy, but invisible for the larger society. Deeper prejudices seeing Roma ethnicity as marginal and as inhabiting a different cultural world serve to keep Roma children at the periphery of social concern.

Following the argument developed by M. Liebel (2004), Roma may risk being labelled as culturally backward in comparison with “developed societies”, or legally culpable since they involve children in what appears to be oppressive practices. This research argues that Roma children’s work is to be read as “the best solution at hand.” People are pragmatic and tend to solve their own problems in their own way, even if the solution may be perceived as “inadequate” by other social groups.

Under these circumstances, for the majority of Roma children, a discourse advocating children’s rights to work may be more precarious than promising. Ultimately, it may become a powerful political tool for delegating the responsibility of what appears to be a social problem to other “accountable” agents: Roma communities, families and children themselves.