Understanding the role of structural factors and realities in normalizing child labour in urban slums of Bangladesh

Abstract Child labour remains widespread in the urban slums of Bangladesh. Empirical studies indicate that various local-level factors drive poor families and children to engage in child labour. However, the role of structural factors and environmental realities is underrepresented in the current scholarship. This investigation examined the role of these factors in normalizing child labour in the slum communities of Dhaka. The researcher adapted a socio-ecological model to develop a conceptual framework for collecting qualitative data from the slum communities of two recently urbanized areas of the city: Hazaribagh and Matuail. A five-month fieldwork project was carried out to conduct 40 semi-structured interviews and two focus group discussions with parents and community members. The data analysis resulted in the following findings. A lack of access to schooling pushed many children into labour. The opportunities for children’s employment in the informal sector remained abundant, and the absence of documents complicated law enforcement. Due to inconsistent adult income, many families are involved in child labour to increase their household income. Local employers chose children over adults for certain occupations due to their low wages and easy-to-control nature. Natural disasters, climate change, and family difficulties drove a large number of families to these slums, and informal employment opportunities for both adults and children drew them there. The shortage of playgrounds and specialized schooling shaped children’s tendency to work more. All together, child labour has become a cultural reality. Policies and interventions aiming to reduce child labour must influence these structural factors and provide support to families and children.

a cultural reality.Policies and interventions aiming to reduce child labour must influence these structural factors and provide support to families and children.

Introduction
Emerging evidence illustrates that the global child labour situation has recently worsened, particularly in post-pandemic reality.According to the latest survey estimate, in 2020, about 160 million children (63 million girls and 97 million boys) were engaged in child labour worldwide (ILO & UNICEF, 2021).About half of those (79 million) boys and girls are engaged in hazardous child labour jeopardizing their health, moral, and safety.Although progress continues among the age group of 12 to 17, the situation has worsened among another age group.In 2020, 16.8 million more children aged 5 to 11 were engaged in child labour than in 2016.The report further noted that by the end of 2022, an additional 8.9 million children would be engaged in child labour because of the pandemic-driven rise in poverty.It means that in 2023, the actual number of child labourers may have been well above 176 million.In contrast, as part of the 8th goal of the Sustainable Development Goals (SDGs), countries are committed to undertaking effective measures to eliminate all forms of child labour by 2025 (Garcia & Galvão, 2021).Over the last few decades, International Labour Organization (ILO) and United Nations Children's Fund (UNICEF)-led global institutions have introduced several international instruments to cease this harmful practice.Governments worldwide have also enacted legislation and implemented policies and programs in support of those global measures (Hoque, 2021b).Nonetheless, child labour reportedly remains not only pervasive but also on the rise in many contexts, especially in Sub-Saharan Africa and South Asia.Why?This puzzle necessitates a comprehensive comprehension of these contexts.What remains unexplored?What are those new factors that contribute to the high prevalence and wide acceptance of child labour?
In four major South Asian countries-Bangladesh, India, Nepal, and Pakistan-24 million children (aged 5-14) are out of school due to labour, and over 15 million children (aged 5-17) are trapped in hazardous labour, producing and supplying products for local and global markets (Zaman et al., 2020).Because of their extremely high prevalence, some industrial places in these countries have been identified as the "hotspots" of child labour (Subrahmanian & Groppo, 2020).Major cities in the region have experienced rapid urbanization over the past few decades, resulting in massive rural-to-urban migration, increased urban poverty, and practices of dangerous child labour in industrial employment (Ensing, 2009;Hoque, 2022;Save the Children, 2012;World Vision, 2014).Many of these locations are situated on the fringes of megacities, where harmful child labour has become entrenched due to various structural realities and environmental factors.In recent times, the COVID-19 pandemic has exacerbated the issue of urban child labour in the region, leading to greater marginalization of these communities (Idris, 2020).Hence, urban child labour has become much more complicated across these contexts (Odonkor, 2020).However, there has been insufficient research conducted on urban informal situations in the South Asian region, as indicated by recent studies (Idris et al., 2020;Yunus, 2020).
Bangladesh is a South Asian country in which child labour is widespread and widely accepted.Dhaka is the fastest-growing capital city in the country.Recent data shows that the city hosts hundreds of child labour hotspots (Maksud et al., 2022;Quattri & Watkins, 2019).Evidence illustrates that child labour remains widely accepted and practiced among the local communities.This research aimed to focus on the environmental realities and structural factors in the normalization of child labour.The question was: how do structural factors and environmental realities influence the prevalence of child labour in Dhaka's slum communities?The structure of this paper is as follows.This introduction section is followed by a literature review section that highlights the justification of this study and the gaps in current academic knowledge.Subsequent sections present the methodology and findings of this research.The concluding section summarizes the paper and lays out directions for relevant researchers and policymakers.

Literature review
Child labour is a contentious issue for which there is no universal definition.Commonly, child labour is defined as labour that robs children of their formative childhood and dignity and harms their physical and mental health (ILO, 2018).ILO's definition of child labour is primarily statistical and includes the following terms: "age of the child," "the number of working hours," "type of work," "time, location, and environment of work," and "what it restricts and offers" (ILO, 2017).Global child labour is a widely attended and researched area of study in the social sciences.Yet, the understanding of child labour remains incomplete and inconclusive due to a few vital reasons.First, although child labour is as old as mankind, the issue has been evolving, which implicates new factors and contexts (Radfar et al., 2018).Second, child labour is a complicated, emotive, and deeply debated issue, and its conceptualizations and understandings are subject to cultural odds (Abebe & Bessell, 2011;Hoque, 2021b).Third, currently active international and national child labour policies and programs have generated a variety of discourses, opinions, and perspectives.Studies suggest that these policies have not adequately considered the voices of local communities, particularly those of families and children directly involved in child labour (Hoque, 2021b;Karikari, 2016).

Recent history of child labour research
Recent research on the reasoning behind child labour can be categorized into three separate waves.The initial wave began in the 1990s.Article 32 of the 1989 United Nations Convention on the Rights of the Child (UNCRC) called for states to protect children from child labour and harmful work (Erdem Türkelli, 2019).Subsequently, in 1992, the ILO established the International Programme on the Elimination of Child Labour (IPEC), aiming to abolish global child labour.Following these developments, several research scholars pointed out the benefits of child labour, and the need to revise the abolition policy.For example, Stadum (1995) argued that because impoverished families benefit from their children's earnings, prohibiting child labour would be detrimental to these families.Based on the significance of work in children's lives, Myers and Boyden (1998) argued that some work that promotes their welfare and growth could be deemed to be in their best interests.Woodhead (1999) brought several working children's voices to the discussion to illustrate that while the majority of children value schooling, they do not see it as an alternative to work.In this period, several researchers theorized the supply and demand of child labour from various socio-economic perspectives (e.g., Akabayashi & Psacharopoulos, 1999;Basu & Van, 1998;Myers, 1999).Based on this new theoretical and empirical knowledge, it became a general understanding that not all work is harmful for children.Eventually, in 1999, the ILO adopted its Worst Forms of Child Labour Convention (C182) to adjust its priority to eliminating the unconditional worst forms of child labour.
The second wave of investigation began around the year 2000.While a few pioneer scholars theorized the worst forms of child labour and exploitative child labour (Dessy & Pallage, 2001, 2005;Rogers & Swinnerton, 2002), others explored various drivers and factors of child labour across a wide range of contexts, especially in Sub-Saharan Africa and South Asia.Winstanley et al. (2002) examined child labour in global supply chains.While Admassie (2002) explained the high prevalence of child labour in Sub-Saharan Africa, Kabeer et al. (2003) assessed working children's educational and income needs in a South Asian context.Schooling was believed to be a wellestablished alternative to child labour.However, drawing on evidence from Bangladesh, Khanam (2008) showed that children could combine schooling and work.Delap (2001) pointed out that cultural factors were equally important as economic ones in explaining child labour.During this period, a few critical perspectives on the definitional and conceptual understanding of global child labour also drew substantive scholarly attention (e.g., Bhukuth (2008) 2008)).Some also assessed various social programs aiming to reduce child labour.Barrientos and DeJong (2006) found cash transfers to be an effective way of reducing child labour.
The latest wave of research in the 2010s constitutes a multidisciplinary approach to understanding child labour.While some researchers continued to explore various contexts and evidence, a few raised some fundamental questions about child labour.For instance, Abebe and Bessell (2011) recognized three central dissertations on child labour: namely the work-free childhood perspective, the socioeconomic perspective, and the political economy perspective.Howard (2014) argued that since child labour remains a reality, it must be regulated, not criminalized.Fors's (2012) review argued that child labour policies must target and influence the underlying factors (i.e., poverty, market flaws, and access to schooling).After recognizing the positive role of work in children's lives, many have argued that work and education cannot be viewed as two separate phenomena (e.g., Aufseeser et al., 2017).Bourdillon and Carothers (2019) advocated for the meaningful participation of local communities in the conceptualization of child labour.The Young Lives cohort study series carried out by the Oxford Department of International Development has brought various perspectives and analyses to understand child work and childhood in the global South.For instance, Morrow and Boyden (2018) analyzed Young Lives longitudinal evidence on children's work from Ethiopia, India, Peru, and Vietnam and made a call to (i) focus on the most harmful work; (ii) implement anti-child labour law sensitively; (iii) tackle care work and family poverty; and (iv) provide working children with access to schooling.
Emerging empirical work has focused more on the worst forms of child labour in newly urbanized and industrialized contexts (Maksud et al., 2022;Quattri & Watkins, 2019).Recently, Brando (2020) asked if there was anything wrong with child labour.In short, the theoretical and empirical research in this wave established that the ILO's age-based child labour abolition policy must be revised; poverty remains an underlying factor; and scholars must consult local communities to explore context-based socio-cultural factors.

Child labour in urban Bangladesh
Bangladesh's post-independence economy was mostly dependent on the agriculture and service sectors (Tama et al., 2021;Sarker et al., 2020).The population was largely rural, and poverty was extensive (Khanam, 2006).Historically, many children in Bangladesh worked in rural agriculture and family farms (Haque et al., 2015;Khanam, 2006).Globalization and trade liberalization accelerated child employment across labour-intensive industries in South Asia (Herath & Sharma, 2007).Bangladesh adopted policies to liberalize its economy in the 1980s and 1990s to attract foreign investments, accelerate industrialization, and diversify exports (GoB, 2003).A few export industries, including ready-made garments, jute, and frozen foods, boomed immediately (Mamun & Hoque, 2022).These urban-based industries employed children to reduce their production costs and increase profits.For instance, White (1996) reported that Bangladesh's garment industries employed about 75,000 people in the early 1990s, and 10 percent of them were below the age of 14. Notably, 70 percent of these working children were girls (White, 1996).Later, a few other formal labour-intensive export industries, such as leather took advantage of low-wage child labour (Ensing, 2009).Industrial deregulation 1 and the privatization of factories encourage private employers to rely on child labour.As Khanam (2006) noted, the number of working children aged 10-14 increased by 4.3 million from 1974 to 2000.While subsistence poverty remained an underlying factor, social norms held child labour to be a cultural practice.Adhering to traditional gender norms, girls tend to work in households while boys mostly serve outside in farms, workshops, and factories (Islam & Hoque, 2022;Tama et al., 2019).Due to its wide normalization, child labour has become a social norm in Bangladesh (Moshiri, 2012).Bangladesh has achieved tremendous economic growth in the last few decades, which has resulted in a rapid reduction of poverty (Hoque & Tama, 2021;Titumir, 2021).However, increasing equality, especially in urban areas remains a great concern (Panday, 2020).With rapid industrial growth, urbanization, and internal migration, the country's socioeconomic landscape has changed rapidly in the last few decades.Many poor and vulnerable families have migrated from rural to urban areas for numerous reasons, including climate change, joblessness, landlessness, extreme poverty, and children's futures (Alam & Hoque, 2022;Majumder & Rahman, 2022).Newly formed industries and their supply chains and informal workplaces created space for cheap labour and child employment (Quattri & Watkins, 2016;Reja, 2017).Numerous families settled in the slum neighborhoods of the newly urbanized industrial areas to meet the demand for low-skilled adult and child labour.For instance, thousands of poor job-seeking families moved to areas in the vicinity of Hazaribagh in Dhaka, responding to the high demand for low-skilled jobs in tanneries and leather factories (Boseley, 2017;Ensing, 2009).Many of these children and adults are processing, producing, and supplying goods for both local and global markets.As a result, child labour became a widely normalized practice in these slum communities (Moshiri, 2012;Norpoth et al., 2014).
Child labour remained widely prevalent in Bangladesh.According to the provisional report of the latest National Child Labour Survey (NCLS) 2022, the estimated number of working children aged 5-17 was 3.54 million, which was 8.9 percent of the total child population (GoB, 2023).The number increased by 0.2 percent in comparison to the 2013 NCLS.The survey adopted the definitions of "working children" and "child labour" based on the principles of the International Conference of Labour Statisticians and the Bangladesh Labour Act of 2006 and found that approximately 1.78 million children (about 77 percent boys and 23 percent girls) were engaged in child labour throughout the country, including 0.44 million in urban areas.About 0.24 million urban children were reportedly working in hazardous conditions.However, various sources provide different statistical data on child labour in Bangladesh.UNICEF adopted the definition provided by the SDGs to carry out the Multiple Indicator Cluster Survey in 2019.The survey reported that 6.8 percent of children (aged 5-17) were in child labour, and the number is much higher among the children not attending school (UNICEF, & BBS, 2019).Another survey conducted in the slum settlements in Dhaka reported that 15 percent of 6-14-year-old children were working in 2019, with average working hours well over the limit (i.e., 42 hours per week) established by national law (Quattri & Watkins, 2019).The study reported that the majority of child participants were out-of-school and employed in the informal ready-made garment sector.About 35 percent experienced extreme fatigue.Two-thirds of children who began working before the age of 10 were found to be unable to read a single word in their native language.
Child labour is associated with children's physical, psychological, and emotional suffering.Ahad et al. (2021) explored the health and educational impacts of child labour on working children in the city of Sylhet, Bangladesh.They discovered that 32.5 percent of child labourers had never attended school.Around 18.75 percent of the child labourers had skin infections, 23.75 percent complained of digestive problems, and 35 percent of the children had suffered physical pain during work.Yet, only 25 percent received medical aid.The workplaces lacked safety and protection measures.In urban areas, many children served in a variety of industries, including transportation, apparel, leather, domestic aid, printing and publishing, waste-picking, chemicals, automobiles, building and construction, hospitality, welding, and furniture (Hoque, 2021a;Kamruzzaman & Hakim, 2018).The findings of a survey carried out in 2019 revealed that 34.6 percent of children (5-17) living in the slum areas of Dhaka were engaged in hazardous child labour (Maksud et al., 2022).Thi et al. (2023) revealed that children working long hours in hazardous conditions are more likely to drop out of school and suffer from psychological and emotional difficulty.In addition, the COVID-19 pandemic has seriously worsened the crisis as many school dropouts engaged in child labour to support their families (Pandit, 2022).Moreover, Hoque (2022) noted that thousands of undocumented children living in the streets of the capital remain vulnerable to various economic exploitations, including child labour.

Factors affecting child labour in urban Bangladesh
A wide range of factors drive the occurrence of child labour in urban Bangladesh.The primary community-level socioeconomic factors that drive urban slum children to work are household poverty, family income requirements, and the non-altruistic nature of their parents (Delap, 2000;Khanam, 2006).Studies point out that parental education and occupation, disability, and family debt directly influence household decisions to engage in child labour (Banu et al., 1998;Jakowan & Afrah, 2019;Nath & Hadi, 2000).Increasing bargaining power of children in families and households explains rising child labour in poor urban neighborhoods (Tariquzzaman & Hossain, 2009).The sexual distribution of work for children also remains explicit.For instance, in the case of domestic care work, of the 10.5 million workers, 90 percent are female and 38.6 percent are below the age of 18 (BILS, 2021).In contrast, more boys than girls are seen involved in harmful child labour in roadside workshops and businesses (Hoque, 2020Quattri & Watkins, 2016).Some notable demographic and context-specific cultural factors contribute to increasing urban child labour.Children's age and place of residence appear to play important roles in determining the type of work they perform in communities across industrial and urban Bangladesh (Salmon, 2005;Tariquzzaman & Hossain, 2009;UCW, 2011).School dropouts are typically employed in occupations with lengthy hours.Family features, including the number of children, birth order, fertility, singleparent households, and whether children live with their biological parents, are also significant determinants (Delap, 2000(Delap, , 2001;;Khanam, 2006).Few studies explored some important cultural factors that explained the growing child labour in Dhaka.Delap's (2001) survey in Dhaka's poor neighborhoods was the first study to evince a few salient cultural factors, including parental attitudes towards children's work and idleness.Parents believed that idleness and sloth were detrimental to boys and could encourage them to engage in criminal behavior.Hard work was viewed as a means of skill acquisition.Ensing (2010) explored the ways working girls' agency was being limited by discriminatory cultural norms in urban communities in Dhaka.In addition to these factors, Tariquzzaman and Hossain (2009) revealed that the lack of social sanctions, parental perceptions about work and low returns from education are key cultural determinants of child labour in Dhaka's slums.
A few quantitative studies have attempted to explore environmental and structural factors influencing child labour among marginalized communities in Dhaka.Two recent surveys of the city's slum settlements have commonly illustrated that the residence of families is an important environmental determinant (Maksud et al., 2022;Quattri & Watkins, 2019).Hoque (2022), while exploring emerging evidence of hazardous child labour in the peri-urban areas of Dhaka, highlights that the environmental ecosystem of the neighborhoods plays a critical role in pushing families and children to engage in child labour.Ali (2021) and Sarker and Islam (2019) mentioned some critical structural factors that influence urban communities to endure child labour.These factors are the lack of adult employment, school dropouts, lack of access to schooling, and rural-urban migration.
As noted above, the current scholarship has mostly focused on the socioeconomic and demographic factors of child labour.Still, emerging evidence affirms the significance of structural factors and realities in understanding the widespread prevalence of child labour in Dhaka, especially in the post-pandemic situation.However, no qualitative research has investigated the role of these structural and environmental factors in explaining the pervasive acceptance of child labour in urban Bangladesh.This gap in the existing literature served as inspiration for this study.

Conceptual framework
This study focused on a few community-level environmental and structural determinants that contribute to child labour.For this purpose, this research adapted a socio-ecological model (derived from socio-ecological systems theory) to identify and examine those factors.The origins of the utilization and implementation of socio-ecological systems theory may be traced back to the pioneering contributions of Kelly (1966) andBronfenbrenner's (1994) seminal work on the ecology of human development.The model is based on Bronfenbrenner's early work (1977,1979,1993) that conceptualized the developing person (the child), the environment, and the interaction between the two (Derksen, 2010).
Initially, Bronfenbrenner contended that the understanding of human development must consider the whole ecological system in which growth takes place.Later, he revised his thoughts to include a child's biological and psychological progression in a social setting.Bronfenbrenner (1994) theorized the ecological models of human development in 1993 into five contextualized and interlinked environmental systems, namely microsystems, mesosystems, exosystems, macrosystems, and chronosystems.The child's immediate physical and social surroundings are referred to as the microsystems, and the interactions between two or more systems containing the developing person (e.g., home, school, and workplace) are referred to as the mesosystems.The concept of exosystems signifies the dynamic interactions that occur between two or more environments, with at least one of these environments not directly involving the person.The macrosystems encompass the broader cultural or subcultural context, encompassing knowledge, worldviews, social norms and beliefs, conventions, risks, materials, and opportunities that shape an individual's life trajectory.The chronosystems encompass various longitudinal characteristics, such as family structure, place of residence, socioeconomic position, employment, and everyday life abilities.Many social scientists and research organizations adapted this model in their social research and interventions and commonly referred to it as the socio-ecological model (Bicchieri, 2016;Caperon et al., 2022;Henderson & Baffour, 2015;Kilanowski, 2017;Lee & Park, 2021).A few contemporary studies also applied this model to investigate and analyze factors of child labour in their empirical research across various contexts (Finigan-Carr et al., 2019;Liao &Sung Hong, 2011;Perez, 2021;Takyi, 2014).
In a community-level research context, various factors play out at different socio-ecological levels (or systems), individual to societal.Assumably, these factors altogether make child labour a widespread phenomenon in the community.However, this research mainly focused on the structural and environmental factors that operate through the four socio-ecological systems of the community concerning child labour.Figure 1, which presents the conceptual framework of this study, describes the community-level structural and environmental factors in the box placed on its right side.The structural factors are access to schooling, adult employment, migration, and employment opportunities for children, while the environmental factors are the material and locational realities that working children and their families live in.As underscored in the literature review section, these factors are important to understand. Figure 1 outlines the four main environmental systems, namely the individual (which in this case is a child), the family, the community, and the societal structure.Many other environmental factors may influence these systems, but this research investigated the locational and material realities that directly shape community behavior in relation to child labour.

Data and method
This study involved a natural context and a real-life issue (i.e., child labour) to provide an in-depth understanding.While this study purposefully adopted a qualitative approach to answering the research question, constructivism motivated its research design.Constructivism is a philosophical paradigm that prioritizes, from an ontological standpoint, how an individual constructs their own conceptions of reality through reasoning (Lincoln & Guba, 1985).The constructivists believe that knowledge consists of mental structures that are surrounded by relative agreements (Basnet & Hoque, 2023;Žukauskas et al., 2018).To construct knowledge with the local community, data were collected from two locations in Dhaka city.The description of the study areas and data collection methods follows.

Study areas
The two locations where this research was carried out are Hazaribagh and Matuail.As Figure 2 points out (see two black squares on the city map and respective area maps on the right side of the image), these two areas are located on the outskirts of the capital, where various manufacturing facilities are located.
Hazaribagh, located in the southern part of the city, is historically known for its toxic and polluted environment filled with solid waste from tanneries and leather industries (Biswas & Rahman, 2013).A 2014 survey of slums in Dhaka reported that there were 243 slums in Hazaribagh (BBS, 2015).The tannery facilities have already been relocated from Hazaribagh to outside Dhaka (Mendoza & Islam, 2018).However, many tanneries and skin processing factories continue to operate informally in Hazaribagh and engage a large number of children in child labour (Gorman, 2021;Zaman, 2021).In fact, children and families from slums and neighborhoods are engaged in harmful child labour not just in tanneries but in several other sectors.Child labour remains a widespread practice in the area.The other location, Matuail, is situated in the peri-urban areas where many printing and press factories, waste processing facilities, waste landfills, electronic and plastic workshops, and wire houses are located (Khan & Sultana, 2017).Reportedly, many working children serve in these facilities, especially as waste pickers and in the printing and paper industries (Alam et al., 2021;Maksud et al., 2022).In search of jobs and cheap housing, many poor families have recently moved into Matuail's slums from rural areas and other places in Dhaka 2 .Both areas are known for hosting thousands of families engaged in child labour in a wide range of informal sectors.

Data collection methods
A prior review of secondary sources informed and guided this investigation.A six-month fieldwork project was carried out in two locations from September 2022 to January 2023.The first month was spent creating rapport with local community leaders through two non-government organizations (NGOs) 3 based in Dhaka.In the following five months, two research methods were used to collect qualitative data: semi-structured interviews and focus group discussions (FGD).Qualitative interviews are the most effective method to understand another person's viewpoint (Mahat-Shamir et al., 2021).Semi-structured interviews allow a researcher to discover the deep thoughts of the participants and follow their trajectories as the dialogue unfolds (Magaldi & Berler, 2020).The FGD helps the researcher understand how participants' views interact to construct conflicting and collaborative comprehension (Gibbs, 1997;Longhurst, 2003).
In each location, a semi-structured questionnaire guide 4 was used to interview the key informants: 12 parents and eight community-level actors (i.e., opinion leaders, schoolteachers, employers, and elected representatives).The open-ended questions were structured to extract insights, opinions, and nuances.Vignettes 5 were used to explore participants' views about reality.A hypothetical adolescent character named Razu was used to illustrate a series of scenarios to the participants via vignettes.For example, 'Razu, a 15-year-old boy, lives with his poor parents here in this slum.He aspires to work to support his family.What kind of work would Razu find here, and how?' The questionnaire was tested through a few trials with randomly selected community members.Subsequently, the selection of participants was performed using the snowball technique.The section conditions were that the participant (a) had been living in a slum settlement in one of the locations for a minimum of two years, (b) was over 18 and able and willing to grant consent to participate in the research, and (c) was aware of child labour issues in the community.Each interview lasted about an hour.The same procedure was followed for the two FGDs.Each FGD consisted of five to six participants and lasted approximately 90 minutes at each location.Both males and females were included.The total number of participants was 51 (40 in interviews and 11 in FGDs).Table 1 summarizes the data collection methods and participants' details.
In addition, the researcher maintained a diary to record the field notes.The notes were used to complement the findings in relation to answering the research question.

Consideration of ethics and limitations
The researcher was aware of the ethical matters and implications at all stages of this research work, from data collection planning to report writing.The ethical application for conducting the field research was reviewed and approved by Coventry University's (UK) ethical review board.The project reference number was P140699.The researcher followed the instructions and conditions attached to the application and adhered to the laws and regulations of Bangladesh in relation to data collection and research.However, readers must be aware of a few critical limitations of this research.First, the findings of this research must be referred to in a specific context.A generalization to a wider context and reality may not be appropriate.Second, child labour is generally viewed as undesirable in advanced societies.Therefore, a few participants may have been influenced by social desirability bias.Inclusion of working children in the research could bolster the study's findings.Due to ethical constraints, it was not feasible.

Findings and analyses
The collected data were initially anonymized for analysis.Subsequently, the data were translated into English with care so that no meanings were lost.The conceptual framework of this study guided the thematic analysis.The framework served as the basis for data analysis, encompassing key themes such as schooling, employment, migration, and the environment.Additionally, new issues that emerged throughout the study, such as the allocation of time and adult employment, were also incorporated.The data analysis technique was reiterative.For writing and analysis purposes, interviews were coded (see Table 1).The code "P" refers to parents, and "C" refers to community members.The first twenty (1-20) referred to the participants in Hazaribagh, while the rest (21-40) were in Matuail.In the subsequent presenting the findings and analyses, the role of ambient reality and structural factors in the normalization of child labour in the two investigated Dhaka communities is explored.

Lack of access to schooling pushes children into labour
Access to education is a vital factor associated with the occurrence of child labour in marginalized communities.Multiple government initiatives in Bangladesh seek to provide poor and disadvantaged children with access to education.The government-run primary schools charge no tuition fees for students, and the government provides the students with books.While the secondary-level students also receive books, the male students pay tuition fees.The government provides free education along with stipends 6 for female students up to the higher secondary level (Rahman, 2013).However, all parents must bear other educational costs (e.g., transport, private tuition, educational materials, school uniforms), which poor families struggle to bear (Jasim, 2022).Multiple government and nongovernment programs offer non-formal educational opportunities for out-of-school children across rural and urban areas.For instance, the Ministry of Labour and Employment runs a non-formal technical training program for children engaged in hazardous child labour in municipal areas (MoLE, 2020).Several NGOs run non-formal and vocational educational programs for underprivileged children, including those who live in Dhaka's slums (Mollah & Jahan, 2023).Regrettably, during the COVID-19 pandemic, the prolonged suspension of schools caused a sharp rise in dropouts, and many NGO-run non-formal schools paused their services for disadvantaged children, creating uncertainty among poor families (Islam & Hoque, 2022;Touhid, 2021).
The analysis of interview data revealed that 17 of 24 parents from both study locations reported an absence of access to school for working children.Six of the rest shared that a few night-time schools had been offering educational opportunities for working children before they closed during the pandemic.One participant chose to remain silent on this topic.Participants in both FGDs reported similar information.The following testimonies help understand the situation.The community-level actors also shared their thoughts on this matter.The following quotes from an employer and a schoolteacher are noteworthy.
A few adolescent workers in my factory regularly join a portable school that sets up here 2-3 times a week.I am glad that they have the opportunity.[Interview/C7] I teach at a local private school.Any adolescent is welcome to enroll.Working children do not attend our school for two reasons: they can hardly afford it, and even if they could, they must labour during school hours.[Interview/C5] The above statements address three pertinent issues.First, many working children had the intention of educating themselves by combining work and schooling.The reasons they could not attend a school were: little or no availability of specialized schools; work fatigue; inability to bear educational expenses; no set working hours; a long way to school; and inadequate educational support.Second, the COVID-19 pandemic had disrupted children's lives and livelihoods in relation to education.As a result, many dropout working children used the opportunity to work longer hours and earn more to support their families.Third, while subsistence needs forced some children to prioritize child labour over schooling, some children just did not want to go to school.Altogether, neither parents, community members, nor employers see child labour as fully unacceptable.
These findings reaffirm that poverty was an underlying factor.Nonetheless, the lack of adequate educational support and access led many families and children to engage in child labour in both communities.This made child labour (combining with or without schooling) a widely accepted community practice.

Abundant employment opportunities are the most alluring factor
Access to and opportunity for paid work for children was identified as the biggest pulling factor for child labour in both communities.As mentioned in the literature review section, children (especially adolescents) in Dhaka city are engaged in a wide range of economic activities.However, some work activities are context-and location-specific.In the case of Hazaribagh, many adolescents were involved in the apparel industry while in Matuail, waste picking and recycling work remained widespread.All research participants commonly reported that the opportunities for child employment were plentiful in both areas.Notably, participants did not mention any illegal employment of children in regularized formal sectors.Almost all the jobs mentioned were in the informal sector.Based on participants' information, Table 2 presents the list of work that children usually perform in the areas.
As Table 2 describes, adolescent children performed a variety of activities.Many children also worked in family businesses.While being asked about categorizing work as "beneficial" or "harmful", participants commonly agreed that jobs that do not put children's physical health at risk can be beneficial for them.For instance, a community leader from Hazaribagh shared: A lot of children, especially adolescents, in this area are economically active.Many of them are employed in shoe factories and retail stores.The families of these children have formed slums to live in and work around.Little children are also involved.In my opinion, these jobs are generally beneficial for children since they can support their families with their income.However, I would classify certain occupations as harmful to children, particularly those involving chemicals and an unhealthy or filthy environment.[Interview/C4] Nearly all participants (parents and community members) viewed children's involvement in so-called "beneficial" work as a normal phenomenon, while many of these jobs can be categorized as child labour as evidence suggests these jobs require long working hours, keep children away from schools, and can damage children's mental health (see Maksud et al., 2022;Quattri & Watkins, 2019).Evidently, parents and community members' views regarding harmful work do not match the ILO's conceptualization of child labour.
The data shows that all recruitments were conducted verbally and that age determination was arbitrary.The absence of documents (such as a birth certificate or national identification card) facilitated the procedure for both parties.No party considered the legal implications.Low levels of awareness of national child labour laws existed among parents and employers.Low community awareness regarding adolescents' employment conditions was a prevalent sentiment expressed by FGD participants.They added that they were unaware of any government or nongovernment awareness campaigns.
Two employers from Matuail shared: We circulate the recruitment advertisement through posters and word of mouth.The advertisement says 8-10 a day; overtime may apply.The parents get to know about the recruitment through their social networks.Sometimes parents bring their children, and sometimes children come alone.The recruitment is done verbally; there is no agreement.We recruit based on the apparent physical fitness of an adolescent, not their age.Our adult workers can decide immediately if the boy or girl can learn our trade within a week.Respectable members of this community often recommend adolescents as candidates.[Interview/C15] We have national laws that say children under 15 7 cannot be recruited.But many are working here.Often, their parents do not even know if their children are working.The laws are not implemented.People are not aware of the laws.[Interview/C14] Many statements, including the above two, indicate a highly active child labour economy in both study areas.The high demand for child labourers was met by an adequate supply.

Inconsistent and informal adult employment makes child labour widespread
Several theoretical and empirical studies concluded that working children come from destitute families, and child labour is more prevalent in households where adult members are not optimally employed (Doran, 2013;Galli, 2001;Salmon, 2005).This is because a lack of adult employment leads to income deficiency and household poverty.Such households tend to increase their household income through child labour.
Findings and fieldnotes indicate that adult employment was an important structural factor at the community level.According to interview data, families with working children could be divided into three categories.The first group consists of families in which all adult members were employed but still required child labour to satisfy their financial obligations.It could be due to the large size of the family or the low income of the adult members.In the second category were families that were fractured or vulnerable for a variety of reasons.It could be a solitary parent, a parent with a disability, or a second marriage.The third category of families were those with high family debts.Adult employment was a crucial factor across all three groups.
Participants commonly convey that adult family members of these vulnerable families were mostly employed in irregular and part-time informal jobs in and around the study areas.Their income was inconsistent and unsecured.Several FGD participants also raised this concern.They reported that regular white-collar employment in local factories and offices required credentials (e.g., educational and professional qualifications, certified training, computer, and communication skills) that locals lacked.Therefore, outlanders-including commuters-filled the majority of these positions.Several parents, after reiterating this point, explained that local factories and businesses favored children over adults due to their low wages and easy-to-control nature.They further added that many parents who work in these local factories and enterprises bring their children with them and work together.Due to security concerns, they do not wish to leave their children unattended at home.

Rural to urban migration remains an underlying cause
Decades ago, researchers identified the relationship between family migration and child labour (Hashim, 2007).However, analysis regarding how migration causes child labour has remained limited, especially with internal parental migration (García Andrés et al., 2022;Van De Glind, 2010).As mentioned in the literature review section, emerging evidence clearly indicates that child labour has concentrated in the areas in and around Dhaka where people have migrated from rural areas in recent decades.Bakth and Hasanuzzaman (2022) highlighted the scarcity of literature concerning temporary migration and child labour in Bangladesh.The most significant factor that pulls migrant families to urban slums like those in Hazaribagh and Matuail is the employment and better livelihood opportunities in the informal sector (Hasan, 2019;Sohel, 2017).Unplanned rapid urbanization and industrialization had created a demand for low-skilled labour (Alam & Mamun, 2022).Major push factors include natural disasters, climate change, extreme poverty, loss of land, and family problems.
findings of this research also indicate that child labour is more prevalent among migrants in both study areas.Local participants shared their understanding of the issue.
Families who migrate here from other cities or rural areas have to live in rented houses.They are not immediately hired when they arrive.If they reside here, they must pay for housing, food, and other incidental costs; consequently, the children and adolescents of these families work more.[Interview/P4] Migrants pay a higher cost of living than locals for a variety of reasons (e.g., house rents, establishment costs).Parents often struggle to meet those additional financial obligations.The number of family members in migrant households is also high.These immediate needs drive families engaging in child labour.Once they get involved in child labour business, it is hard to stop.[FGD1] Once they left their village, they had to start here afresh.Many migrant families came here with nothing-no money no property.They had no other way but to earn from whatever they got.[Interview/C15] Often, migrant families bring their relatives too.Many of these relatives are children.When they start earning here, they neither go back nor continue their education here.Most migrant parents have never been to college.[Interview/P23] These statements highlight that some of these slum communities have formed based on informal work opportunities.Child labour is an integral part of this informal livelihood ecosystem.A few issues are critical to understanding the role of migration.First, migration is a continuous process.People who had migrated from villages to urban or peri-urban slum areas of Dhaka continued to influence their relatives (individuals and families) to migrate to Dhaka.The opportunity for children to find employment is an important element of this migration.Second, due to migration, many children could not continue their schooling.Once a child dropped out of school, families had no alternative to child labour.The scope and returns of non-formal education are limited; there is a scarcity of playgrounds, and parents fear of the new and unknown environment (a child getting involved in crime and drugs) is profound.Data indicates that the long suspension of educational institutions during the pandemic accelerated migration.Third, when migrant child workers became adult labourers and parents, they viewed child labour as an accepted and normal phenomenon.Since most slum dwellers are migrants, child labour remains a well-accepted practice of livelihood.

Environmental and material reality creates a favourable condition for child labour
The environment shapes the behavior of individuals and families.Unfortunately, the slum-dwelling communities live in the harsh realities of Hazaribagh and Matuail.Various environmental risks (sound pollution, waste pollution, lack of water and sanitation) and a lack of child-centeredness and safety threaten children's wellbeing (Vostanis et al., 2021).Around 75 percent of these slum households live in one room, and several households often live under one roof (UNICEF, 2018).Children do not tend to stay home.There is hardly any playground for children (Hoque, 2022).Public services are limited.Children often perform their outdoor activities on the street.Since most of these families have migrated to these areas, they do not have access to urban citizenship rights.Moreover, many of these families lack the necessary citizenship documents to receive government support.Due to these conditions, slum dwellers remain impoverished and deprived.
The findings and fieldnotes show how these environmental realities shaped children's timespending attributes in both communities.When asked about how children spend time, all participants provided similar information.Slum-dwelling working children did not like to spend time at home.They spent time with their friends, mostly gathering on or by the streets.After daytime work, most children spent time on their mobile phones, especially on social media platforms.The following statements these findings: Children here work during the daytime.In the evening, they spend time mainly on social media (online); some of them play in open fields and streets.Those with no smartphones spend time chit-chatting and performing outdoor activities with their colleagues and friends.
[Interview/C1] The way children and teenagers spend their time is: they work during the day, they have the evening to finish work, and after that, they hang out with their peers.Those who have smartphones watch YouTube, use Facebook, play sports on holidays, watch TV, and help with household chores.Some of the boys also go on a group tour.[Participants in FGD1] Children here cannot enjoy a good environment.They are forced to spend time playing outdoor games, on Facebook, and on TikTok.Girls who serve as domestic help must work in the evening too.[Interview/C9] Since they [working children] do not go to school regularly, they spend time on Facebook, YouTube, and TikTok.Boys also go on trips on off days.[FGD2] These statements also convey two other important issues.The first one is the sexual division of labour indicating that boys were more likely to work outside while most girls served in private spaces (e.g., domestic help, cleaning).The second one is regarding the agency of working children.Working children, especially boys, seemed to enjoy greater agency in making day-to-day and strategic decisions.Although some parents wanted to send their boys to school, they [the boys] often got influenced by their peers and friends.They chose to drop school and start working long hours.Supporting families and households with income gave them greater freedom.
Although participants did not mention it, secondary sources confirm that many of these working children are also involved in petty crimes and do drugs (Atkinson-Sheppard, 2016;Sharmin & Disha, 2021).A 2010 survey found that about 85 percent of the street children in Dhaka were addicted to drugs (Jahan, 2020).Such involvements forced children to opt for high-paying jobs that required long working hours and additional risks.Besides, as mentioned earlier, parents feared that their children would use their idle time to do something wrong.That is why parents preferred child labour to other activities.Thus, environmental realities create a favorable condition for widespread child labour in these slums.

Conclusions
This research aimed to explore the role of structural and environmental factors in child labour in Dhaka's slum communities.The literature review informed the conceptual framework, which directed the qualitative data collection.Data were collected through interviews and FGDs with slum communities living in Hazaribagh and Matuail.The analysis of the data conveyed several valuable findings.
Firstly, in these communities, limited access to formal and non-formal education was the most significant structural factor that led to child labour.Even though children were compelled by subsistence requirements to choose labour over education, a lack of access to specialized schools, their remote locations, and inadequate support exacerbated the problem.Secondly, opportunities for children's employment remained widespread across both communities.Participants named a wide range of activities that boys and girls could choose to perform.The community member's conceptualization of "beneficial" and "harmful" child work was different than what national and international policies put forward.Child labour primarily occurred in the informal sector, and verbal contracts were the norm.The absence of documents complicates law enforcement and renders age determination arbitrary.Consequently, child labour grew as a part of the community's culture.
Thirdly, findings indicated that parents and adults living in these slums lacked the academic and professional required for white-collar jobs in the formal sector.They were compelled to choose readily accessible informal occupations that were prescribed, temporary, and low-skilled.Due to inconsistent income and an unsecured future, many families engaged in child labour to increase their household income.In addition, local employers chose children over adults for certain occupations due to their low wages and easy-to-control nature.Fourthly, migration remained a decisive underlying factor.Numerous families migrated to these slums, pushed by natural disasters, climate change, and family difficulties, and drawn by informal sector employment opportunities for adults and children in newly industrialized areas of Dhaka.Migrant families often brought their relatives (individuals and families) to serve in similar occupations.This rural-to-urban migration caused a break in children's regular schooling and resulted in widespread child labour.
Finally, environmental and material realities compelled families to engage in child labour.The shortage of playgrounds and specialized schooling shaped children's attributes concerning their allocation of daily time.Reportedly, working children spent their leisure time on social media.Child labourers used to enjoy greater agency in their families and were influenced by their colleagues and friends.They discontinued their education and began to work more.Due to safety concerns, many parents did not want to leave their children at home.Additionally, parents dreaded that idle children would become involved with drugs and crime.Consequently, child labour became a practical option for the families residing in Hazaribagh and Matuail's slums.
These results are essential to comprehending the child labour situation in the study areas and are indicative of a broader context.However, readers must recognize the limitations of the data methodology used in this study.The evidence indicates that child labour is profoundly rooted in Bangladesh's urban slums.Future research may investigate the role of additional social and structural factors identified by this study, such as climate change, cultural norms, and social networks.Some policy initiatives can be undertaken to tackle structural and environmental factors at the urban community level.Educational and protective interventions must go hand in hand.Government and non-government organizations must provide working children with adequate educational (formal and non-formal) opportunities.Poor families must receive adequate social assistance.Community members and employers must be made aware of the current regulations so that they adhere to them and provide safety to child workers.
4. The interview guide was designed to collect three types of driving factors for child labour: namely the sociocultural factors, environmental and structural factors.This paper is based on structural and environmental factors.5. Vignettes are hypothetical short stories presented to the participants with the intention of extracting their views and understanding about a sensitive topic.The use of vignettes is increasing in qualitative studies.For more details, see Gourlay et al. (2014).6.Under the Female Secondary Stipend and Assistance Program (FSSAP), the government provides stipends to those females who achieve satisfactory academic proficiency, and attend 75 percent of school days, and remain unmarried.For more details, see Khandker et al. (2021).7.According to the Bangladesh Labour Act, 2006Act, (amended in 2013)), the minimum age for admission to work is 14.

Figure
Figure 2. Study areas (Illustration: Authors' own) ; Edmonds et al. (Edmonds & International Labour Office, & ILO International Programme on the Elimination of Child Labour,

Table 1 . Summary of research methods and participants Location Method Type of Participants Number and Details Code
Many working children and adolescents are unable to participate in educational despite their wish to do so.Previously, there were opportunities to study here.A night school was providing general education, and a morning one was offering vocational education.The morning school used to provide monthly financial support to regular students.These schools are closed now.The reality is that the work opportunities are close by, but the schools are far away.Children and families tend to harness income sources because of their subsistence needs.[Interview/P3]Children's ability to combine education and work depends primarily on the employer.If the employer operates a small and informal enterprise without a permanent location, it is not possible.Most employers have no set work hours for their employees.In addition, the school is 3-5 km away.[Interview/P15]As a result of the pandemic, there are almost no opportunities for working children and adolescents to combine work and education, as specialized schools and night schools have closed.[Fiveout of 6 participants in FGD1 agreed] There are no specialized or night schools available in Matuail.No one here has so far thought to provide such educational facilities for child labourers.[All5 participants in FGD2 agreed]