A BASELINE SURVEY OF STREET CHILDREN IN BANGLADESH

Submitted to The Bangladesh Bureau of Statistics National Child-Labour Survey, 2002 –2003 Parishakhan Bhaban, 4th Floor, (Block –1) E-27/A, Agargaon, Dhaka.

by

Foundation for Research on Educational Planning and Development (FREPD) Sonargaon Road, Palashi Crossing, Dhaka University Campus, Dhaka. e-mail : frepd@dhaka .net, Phone : 8610556

December 7, 2003

Acknowledgement

In order to combat the situation of child labour in its worst forms effectively, the size of the affected population must be known. The number of street children in Bangladesh is quite big and until recently the incidence and distribution of street children at both national and regional levels was not known. Their reality in terms of their working, living and health conditions, as well as the degree of hazardousness faced by the street children was also not known. Gathering data on the above issues is very important for planning interventions that would help to know the real situation of street children in Bangladesh. Thus the BBS and the ILO decided to conduct a Base Line Survey of Street Children with the main objective of the survey being to attempt to quantify the number of the street children population at the national level in Bangladesh. This data-collection exercise would generate benchmark data covering different aspects of the realities of Street Children in Bangladesh. BBS and ILO entrusted the FREPD with the responsibility of conducting this important survey. The FREPD thanks the BBS and ILO for selecting FREPD for the conduct of the survey. I am greatful to A.K.M. Musa; the D.G. of BBS for his generous support. The study team thanks Mrs. Sanjukta Mukherjee of ILO and Mr. Zobdul Huq of BBS for conceptualizing the study and for providing technical support at all stages of the survey, including survey design, developing the data collection instruments and editing the draft report. Their technical comments on the tables and draft report were very useful in improving the quality of the presentation of the report. The research team thanks Mr. Atin for his cooperation. The research team is grateful to all members present in the workshop, particularly to those who gave feedback on the report on the Street Children. The study team is also greateful to Mr. H.S. Sujeewa Fonseka, Chief Technical Advisor, US-DOL Project, and Mr. Gopal Bhattacharya, Director, ILO Area Office, Dhaka for their support in conducting the survey. The FREPD wishes to express its gratitude to the law enforcement agency, the key Informants and local people of the sample towns for extending their cooperation to the research team and field staff. I wish to record with appreciation the sincere service and hard work of the field supervisors and data collectors. I also thank Mrs. Koheli Parveen and Mr. Satter for their secretarial services. I wish to acknowledge with deep appreciation the cooperation of Prof. Obaedul Huq, Mrs. Zahurun Nessa, Mr. Zahid and Mr. Nitail Chandra Paul for helping the team.

(Dr. Kazi Saleh Ahmed) Team Leader and President, FREPD.

ii

Study Team

Dr. Kazi Saleh Ahmed (Team Leader) Dr. M. Mosleh Uddin Mrs. Shamima Islam Mr. M. Nazmul Huq Mrs. Shamsun Nehar Mrs. Zahurun Nessa Prof. Obaedu Huq (Coordinator)

iii

Acronyms

BBS BCW BS BSS CBO CW FGD GOB GO FPBS IPEC ILO KI KII LGB M&E MLE MSW MWCA NCLS NGO QCS RQCS SC SF SFBS SU TBR TC TOR TPBS UP UNICEF WFCL

= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =

Bangladesh Bureau of Statistics Baseline of Child Workers Baseline Survey. Baseline Survey Specialist. Community Based Organization. Child Worker. Focus Group Discussions. Government of Bangladesh. Government Organizations. Financial Proposal for Baseline Survey. International Programme on the Elimination of Child-Labour. International Labour Organization. Key Informants. Key Informants Interviews. Local Government Bodies. Monitoring and Evaluation. Ministry of Labour and Employment. Ministry of Social Welfare. Ministry of Women and Children Affairs. National Child Labour Survey. Non-Government Organizations. Quick Count Survey. Result of Quick Count Survey. Street Children. Sampling Frame. Sampling Frame for Baseline Survey. Sampling Universe. Tender for Baseline Survey. Technical Committee. Terms of Reference. Technical Proposal for Baseline Survey. Union Parisad. United Nations Children’s Emergency Fund. Worst Form of Child Labour.

iv

Content

Page Executive Summary CHAPTER – 1 Introduction, Objectives, Methodology and Data Collection 1.1. 1.2. 1.3. 1.4. 1.5. 1.6. 1.7. 1.8. 1.9. 1.10. Introduction Situation of working children in Bangladesh: An overview Background of the survey Objectives of the baseline survey of street children Definition of important terms used in the survey Methodology for survey Preparation for the implementation of the field survey Field survey Quality control mechanism and reliability of estimates Field realitites and limitations of the data ix 1 1 1 1 2 2 3 4 6 7 7 7

CHAPTER – 2 National Estimates of Total Street Children and Their Background Characteristics 2.1. 2.2. 2.3. 2.4. 2.5. 2.6. 2.7. Estimate of street children by sex and by division Estimate of street children by regions Home regions of street children Demographic characteristics of street children Personal information about the street children Duration of stay in the town Visit to the parents

9

9 9 9 9 10 10 11 11

v

2.8. 2.9. 2.10. 2.11. 2.12. 2.13.

Educational and training issues Reading and writing skill Schooling status of street children (ever attended school) Education Perception about importance of education Skill training and job preference

11 12 12 12 13 14

CHAPTER – 3 Past Work History of Street Children 3.1. 3.2. 3.3. 3.4. 3.5. First job of the street children Age fo the children when first started working Reasons for starting the first job Who put them to work in the first job Whether the children stopped school to work in their first job

15 15 15 15 15 15 16

CHAPTER – IV Current Work, Work Conditions and Related Issues 4.1. 4.2. 4.3. 4.4. 4.5. 4.6. 4.7. 4.8. 4.9. 4.10. Current work Age when started working as street children Duration of street based work Daily working hours Weekly working days Weekly earning, spending, saving and borrowing Busy and slack period Liking or disliking the current work Employer and work condition Company at work

17 17 17 17 17 17 18 18 19 19 19 19 vi

4.11. 4.12.

Arrest by police Mobility of street children within the town

19 20

CHAPTER – 5 Health Hazards and Related Issues 5.1. 5.2. 5.3. 5.4. Sickness Injury at work Protection used during work Smoking habit

21 21 21 21 22 22

CHAPTER – 6 Living Conditions and Social Support 6.1. 6.2. 6.3. 6.4. 6.5. 6.6. 6.7. 6.8. 6.9. 6.10. Sleeping place, type and company Seasonal variation Reasons for searching new sleeping place Payment for sleeping Bed for sleeping Access to bath Type of toilet used Food intake: Frequency, place, starvation Winter clothing Organizational assistance

23 23 23 23 23 23 23 24 24 24 24 24

CHAPTER – 7 Lessons Learned and Policy Implications

25 25

vii

Annex 1: Tables Annex 2: Review of Quick Count Survey (QCS) Annex 3: National Child Labour Survey, 2002-2003 Annex 4: Employers Survey Questionnaire Annex 5 : Matrix Showing Number of Location, Children, their Occupation and Anti-Social Activities

26 58 61 78

79

viii

Executive Summary

1. BBS conducted Quick Count Survey (QCS) of Street Children in November-December 2002. The QCS estimate was 2,573 Street children in Bangladesh. In order to have accurate and verifiable data base for the street children BBS-ILO decided to conduct a base line sample survey of the street children of Bangladesh and entrusted FREPD to carry out the study. FREPD conducted the Baseline Survey in April-May, 2003. 2. A stratified two-stage sampling design has been used to capture the relevant information. A sample of 450 street children were covered following proportional allocation scheme. The sample has been drawn from 17 greater districts (region). Key informants and a few employers were also interviewed. 3. In Bangladesh the situation of street children is hazardous and precarious. There is sufficient evidence that working children live in severe poverty condition and their numbers are increasing. 4. Among all street children surveyed, 97.56 percent are boys and 2.44 percent are girls. An overwhelming 55 percent of street children were found in Dhaka. 5. The poorest children of different parts of the country migrated to the town, particularly to the metropolitan cities: Dhaka, Chittagong and Sylhet. Highest percent of street children are born in, in order of ranks: Barisal, Comilla, Dhaka, Faridpur, Sylhet, Mymensingh, Kishoreganj, Patuakhali, Noakhali, Rangpur, Chittagong Districts. 6. Among the street children covered in the survey, 48.7 percent were of age 11-14 years, 26 percent were of age 14-17 and 25 percent of age 5-11 years. Girls were relatively younger (average age 11.5 years) compared to boys (average age 12.2 years). 7. Parents' occupation reveals that street children come from poorest families having day labor and small business as the main source of income. The parents are predominately illiterate (70 percent father and 76 percent mothers are illiterate). 8. Among all street children, 33.78 percent children's first job was street related job while 66.22 percent had other jobs before they started to be street children. While the list of “first job” is quite long, domestic servant, agricultural worker, begging, cooli, minti, paper pickers, flower sellers feature prominently as being typical of the “first job”. 9. The mean age at starting the first job was 7.81 years, and the mean age of street children at the time of the survey was 12.16 years. This indicates that on average, one street child has been working 4.35 years prior to the date of interview. Poverty was the main reason, and abuse by family members was the second important reason for children joining the first job. 10. As high as 82 percent children stopped school in order to start work while 18 percent combined school and work together.

ix

11. The street children are involved in a wide variety of activities. The most frequently mentioned activities are: collect old papers, cooli/minti, tokai, begging, helper, hawker, shoepolishers, seller of flowers, etc. A small percentage of street children are, as reported by key Informants, involved in theft, snatching, pick pocketing, sex work, drug business, Informer etc. 12. Age wise variation is also noticeable. Begging was predominant among younger children while Cooli and Paper Pickers are common activities among elder children. However, cooli/minti and paper pickers are the most common activities among all children, regardless of age. 13. Daily working hours ranges between 5 to over 12 hours in a day. The modal group is 8-12 hours. The estimated average working hours was 10 hours in a day. It seems that intensity of work was not correlated with age. Thus, age wise variation was small. About 83 percent work 7 days a week and 17 percent work 6 days or less in a week. The average working days was estimated to be 6.77 days. 14. Average weekly earning ranges from as low as Taka∗ 200 to more than Taka 1000. The average weekly earning was estimated to be Taka 287.6. The income is strongly positively correlated with age. 15. More than 57 percent street children reported that they felt sick. The most common sickness, in order of frequency of respondents, were fever, water borne diseases and headache. 16. About 73 percent street children consider their present sleeping place as their permanent place of sleeping. Those who do not have permanent sleeping place generally sleep in bus station, market place or on the street. About 61.9 percent street children sleep with other children while about 33 percent sleep alone. During rainy and winter season the street children sleep at the railway station, steamer and bus terminals. The reasons for changing sleeping place are: night guard’s objection, police harassment, and not safe. 42 Percent street children do not use any thing as bed, 25 percent use jute cloth, and only 16 percent use bedsheet. 17. 60 percent street children take bath every day, 30 percent every two days, and others take bath irregularly once or twice a week. 18. Public toilet and open space are the places where children usually defecate, 42 percent use public toilet, 35 percent open space. 19. 57 percent street children eat three times a day, 39 percent eat two-times a day, and 4 percent less than two meals a day. 88 percent street children eat in the street shop, 5 percent beg food from hotels, restaurants, less than one percent collect food from dustbin. The team found 2 percent street children without full meal during the last 24 hours prior to interview. At least 60 percent children could not take their night meal for at least one night of the previous week. 20. 84 percent children do not have winter clothing.

∗

One US$ = 59 Bangladeshi Taka

x

21. Only 8 percent children are aware of organizations which provide assistance to street children. Only 6 percent visited such organizations. Children who visited the organizations mentioned that they visited for education and food and clothing. Children who knew these organizations but never visited these organizations reported the following reasons: Not sure of getting assistance, difficult road communication and high travel cost. 22. 61 percent street children never attended any formal and /or non formal schools, 30 percent attended formal schools and 12 percent non-formal schools. About 3 percent attended both formal and non formal schools. However, at the time of the survey only 8 percent children were attending schools. Nearly 10 percent children stopped school between when they first joined work and the survey time. 39.3 percent street children ever attended schools. However, among them 22.6 percent just attended class I without completing, 29 percent completed class I, 23 percent class II, 13 percent class III, 6.7 percent class IV and 6 percent at least class V. The average years of education is estimated at 0.672 years with standard error of 0.058. This implies that the overwhelming majority of street children had virtually no education. Poverty and ignorance about the importance of education were the main causes for not going to school. At the time of the survey only 8.4 percent street children were attending schools. Those who are attending schools combine school education and work together. They mostly go to NGO run schools. Poverty and long hours of work in the current job are the important reasons for currently not going to school. The street children, as high as 80 percent, give importance to education. These children are ready to start education if opportunity is created for them. 23. The children expressed their desire to avail skill training if they get any opportunity to have it. The most preferred jobs are driver, mechanics and garage workers. 24. The study revealed that poverty has driven these children to the street for survival. The Government and NGOs should come forward to rehabilitate these children and initiate education and vocational training programmes so that they are able to have a better future.

xi

Chapter – 1 Introduction, Objectives, Methodology and Data Collection
1.1. Introduction

Today’s children are the future of to-morrow. Since they create the world of to-morrow, they are at the heart of social development. The future depends on how children prepare themselves to enter into the world of work. Children who are healthy, well-fed and educated grow up to be productive, innovative workers and responsible adults. But to-day more than one billion children (First Call for Children, A UNICEF Quarterly, 1995 / No1) are condemned to poverty, without jobs, without basic necessities, without hope. According to IPEC, 1994, about 200-300 million children work through out the world, work long hours in poor and dangerous conditions; the hazard that working children face are many and varied. Poverty is the root cause for the incidence and prevalence of child labour. Child labour occurs at exploitative circumstances putting the child’s life in hazardous condition leading to child abuses. Such circumstances impede physical, mental and moral development of the child. Child abuse is closely related to the social development of a country, and highest in countries where incidence of poverty is maximum. 1.2. Situation of working children in Bangladesh: An overview

The absolute and relative size of the population of children in Bangladesh is quite big as a share of the national population. The estimated total population in Bangladesh is 130 million (2001). Among them about 42 million (32.2% of total population) are 5-17 years old. The distribution of population and children are given in Table 1. According to the labour force survey conducted by BBS, 5.8 million children aged 10-14 years were working in Bangladesh in 1990 –91 and this constituted 11.3 percent of the labour force. All the studies conducted so far show that working children live in severe poverty and the number shows an increasing trend. Surveillance data gathered by UNICEF in 1995 show that one million labours are employed in garments industries of whom about 90 percent were female and 1 percent were children below age 14 years. In rural Bangladesh, children traditionally worked on land. However they had the conventional protection of the families. Today’s children are more vulnerable in the urban areas, in informal work sectors, where neither the family nor the law accord protection. On the contrary the employers have vested interests in engaging children, since their labour is cheapest, their working hours can be longest and their bargaining power is non existent. The education of children for long-term life skills has always been underrated for economic gains both by the employers and parents. A substantial percent of child labourers work minimum 9 hours to as long as 18 hours (on average 10 hours a day). About 70 percent of the child labourers do not attend schools, 30 percent get education in addition to their jobs. Of these who are not attending schools, 48 percent gave economic constraints as the reasons. About 68 percent of the children not attending school expressed interest in acquiring education. A study conducted by a donor funded team (Blanchet), depicted a gloomy picture of children’s rights especially of girls in Bangladesh. About child labour, the study cited that most of the labouring children themselves do not mind having to work. What they object to are the humiliation, scorn and the various abuses

they have to endure from their employers and clients. The study revealed “Girls in particular are denied of right for a wage. National statistics show their presence in the labour force to be 10 times lower than the boys. This does not reflect the real situation. Girls are massively present in domestic service and commercial sex work.” However, very poor families were often forced to send their children to work for others. “Misplaced childhood”, a study of the RED BARNET, Danish Save the Children revealed that street children are involved in the following work: * * * * Street sex workers Transport labours Workers in informal sectors Tokai * * * * Occasional workers at hotels, restaurants etc Coolies Rickshaw Pullers/ Van-driver etc Hawkers and others

The Government, NGOs and donor agencies have been concerned over the rapid growth of the child workers and they are looking to find ways and means to gradually and progressively eliminate child labour in Bangladesh. 1.3. Background of the survey

The International Labour Organization (ILO) Conventions No. 138 and 182, define hazardous work by children as any activity or occupation, which by its nature or type has or leads to adverse affects on the child’s safety, health (physical or mental), and moral development. Hazards could also derive from excessive workload, physical conditions of work, work intensity in terms of duration of work where the activity is known to be non-hazardous or ‘safe’. The list of such work is usually to be determined at the national level after tripartite consultation. Thus, the Technical Committee on National Child Labour Survey, 2002-2003, in Bangladesh, ranked the following 5 forms of child labour as being the most hazardous sectors in Bangladesh, and thus urgently needing serious attention. There were essentially two criteria to determine the ranking of the hazardous sectors. The criteria used were: (a) degree of hazardousness and (b) the intensity of Child laborers engaged in these Worst Froms of the Child Labour. The five sectors that were determined on the above criteria to have priority were: street children, child workers in battery re-charging, automobile sector, transport sector, and welding sector. The objective of the Baseline Survey was to measure the incidence and the distribution of street children at the national level in Bangladesh. The degree of hazardousness faced by the street children is not known, so data collected on various modules such as the working conditions, work hours, health, nature of living, gender and age disaggregation, access to services and other socio-economic variables would provide insight into the realities of street children’s lives in Bangladesh. 1.4. Objectives of the baseline survey of street children

The main objective of the baseline survey for the street children in Bangladesh was to establish accurate and verifiable data set for the street children (as defined in section 2 (b) of TOR) of Bangladesh. The specific objectives were to determine: •

the estimate of street children in Bangladesh as a whole, by region, by division; 2

• • • • • • • • • 1.5.

the incidence and the distribution of the street children, by age, gender, socio-economic variables, by type of activity and also by number of activities; the degree of hazardousness for each type of street children; the working conditions of the street children ; working hours, sleeping hours, time spent in other activities including recreation if any; health status of the street children; access to educational, health, social, recreational facilities; their perception regarding child’s right; their knowledge about jobs and job markets; Their thinking about their own future. Definition of important terms used in the survey

Street Children: For the purposes of this survey, Street Children refers to those children aged 517 years who are living (i.e. sleeping, eating and working) on the street of a particular city, town or thana head quarters. These children are floating in nature and they may live in one place for sometime and then move to other places. They are detached from their family or parents. They usually sleep at night on the roadside, railway stations, bus stations, park, and abandoned houses and public places. These distressed children are primarily seen in the big cities. In this study, street children do not include those children who live with their parents/ family on the roadside/ pavements, or in a slum. Division: Bangladesh is divided into six administrative divisions. These are Barisal, Chittagong, Dhaka, Khulna, Rajshahi, and Sylhet. District: The administrative divisions are further disaggregated into districts or Zila in Bengali. At present there are 64 districts. Upazila (Thana): Each district or zila is further divided into smaller sub-units known as upazila. Several upazilas therefore comprise a district (Zila) Region (Greater district): Previously Bangladesh had 23 districts, currently 64 districts. These 23 districts are known as greater districts. Bangladesh Bureau of Statistics (BBS) refers to these greater districts as region. In this survey region refers to greater district. Maid: In this survey maid means a female servant or a male servant, working in other house. Housewife: Housewife refers to a woman who is not employed, self-employed or unemployed but does all household work. Brokers: Persons who act as agents/intermediaries, who assure good jobs to the children, bring them to the town and involve them in hazardous activities. They take away the lion share of the income of the children. Tokai: It is a popular Bengali term used for child ragpickers: A child who moves around the town to pick up various used items like papers, bottles, shoes, cloths, etc is called tokai. Cooli (Kooli): Persons who earn their livelihood by carrying baggages of others in the railway station, ferry terminal, bus terminal are known as Cooli. They are generally registered by the authority. 3

Minti: They do the same job as that of a cooli. But they are not registered. They work in the markets, shops, railway station, ferry ghat, bus station etc. In places where both cooli and Minti work, the latter has lower wage rate. Pir-Fakir, Tantric: They are religious persons who adopt different treatment methods having linkage with religious beliefs. Poor people believe that these Pir-Fakir and tantric have the power to do the miracle and cure the sufferers. Unan–Ayurbedic: Alternative Treatment and medicine based on natural materials such as leaves of trees, minerals, grass, etc. Pre-primary: Education prior to primary class I is known as pre-primary. This is also known as baby class. 1.6. Methodology for survey

The survey of street children was done in a two-stage process. Given that there were so many unknowns to deal with, the first step was to use the estimated number of street children obtained in Quick Count Survey (QCS) all over Bangladesh in 20 regions. The results of the QCS enumeration revealed that in 6 regions (Khagrachari, Bandarban, Rangamati Jamalpour, Kustia and Tangail) the incidence of street children was insignificant. Thus, these regions with negligible incidence of street children were dropped from the sampling frame, thus restricting the sample universe to 17 regions where there was positive incidence of street children. In the second step, a stratified sample was conceptualized. Regions were selected to be the strata with thana/upazilla as the PSU for each strata. The primary unit of observation/measurement was the street child and the selection of children within the selected PSU, was done on a simple random sampling basis. The sample size of 450 was allocated to the strata (Table 4) in proportion to the number of children estimated in the QCS 2002 (Table 2 and Table 3). The unit of measurement was the street child who was selected using random procedure. This provided a basis to estimate the parameters as well as standard error of the estimate. 1.6.1. Number of thanas having street children in QCS in 2002.

The total number of street children was estimated to be 2,573 in Bangladesh, via the QCS (Annex 2). Regional variation was apparent. Data of the Annex 2 shows that in 8 thanas of Khagrachari, in 7 thanas of Bandarban, and 13 thanas of Kushtia, in all 28 thanas of the three regions, there were no street children Comparing column 3 (number of thanas) with column 4 (number of locations) of Annex 2 we observed the following: (a) (b) (c) (d) (e) (f) (g) In Mymensing at least 20 thanas had no street children In Jamalpur ’’ ’’ 4 ’’ ’’ ’’ ’’ ’’ In Kishorgonj ’’ ’’ 3 ’’ ’’ ’’ ’’ ’’ In Tangail ’’ ’’ 6 ’’ ’’ ’’ ’’ ’’ In Faridpur ’’ ’’ 21 ’’ ’’ ’’ ’’ ’’ In Rangamati ’’ ’’ 7 ’’ ’’ ’’ ’’ ’’ In Comilla ’’ ’’ 24 ’’ ’’ ’’ ’’ ’’ 4

(h) (i) (j) (k) (l) (m) (n) (o) (p) (q) (r)

In Noakhali In Sylhet In Rajshahi In In Bogra In Rangpur In Dinajpur In Pabna In Khulna In Jessore In Barisal In Patuakhali

’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’

6 9 33 8 8 20 16 12 14 16 4

’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’

’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’

’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’

’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’ ’’

From the QCS it was found that 259 thanas out of 417 thanas, had no street children. This means that the maximum number of the thanas that would have a positive incidence of street children could be 158 thanas. In fact, during the survey it was found that the number of thanas having street children was much below 158. Based on the assumption of 158 thanas we worked out the minimum average number of street children per thana. This comes out to be 16 children and the average of minimum location/thana comes at 2.6 location. Determination of the total sample thanas The total sample for street children for the Baseline survey has been initially fixed at 400 street children. In our study, we intended to include as many thanas as possible to give coverage of various geographical location within a given strata. If the sample size in a thana was very small, then it would not be cost ineffective. More over it would pose problems in administering the surveys. The use of proportional allocation provided 53 thanas to be selected for the survey. In order to ensure quality data and minimize non-sampling error, it was decided to have on average a sample of atleast 8 children in each thana. This led to an increase in the sample size from 400 to 438 (Table 4) which further increased to 450 to cover non-response (Table 5). The total sample size for the thanas came to 53. Allocation of sample thanas and the sample children is shown in Table 5.

Allocation of 53 thanas and street children to the 17 strata Selection of sample thanas in a given strata It is found during the survey that the number of thanas having street children was much lower than the number estimated during the QCS in 2002 The QCS 2002 identified seventeen regions as having street children. The list of the thanas having street children for the 17 regions is available at BBS. For a given region, let the number of thanas be.k and the street children be m1, m2, ... mk for k thanas arranged as follows:

5

Thana no. : Street Children Probability of selection

1 m1 P1

2 m2 P2

K mk Pk

Total Σ mi=mo Where Pi = mi/mo

Thanas were selected using the above scheme. Thanas having less than 8 children were merged with the neighbouring thanas before selection. The same procedure was adopted for all the 17 strata. Estimation of parameters The procedure for estimating a particular parameter is adopted following the sampling ˆ design: Most parameters intended to be estimated are proportions. P is the parameter and p is the estimate.

The formula used for estimation was ai Ni ˆ pi = , Wi = ni N

N = Total number of Street Children Ni= Total number of Street Children in ith stratum

ni = nWi = n ai N ˆ Pi = n Ni
ˆ ˆ P = ∑ WiPi

Ni N
Where

ai

Number of units having the attribute

∑ Wi ni ∑ N . n . Ni
∑ ai
Ni ai N Where ai = number having the attributes

ai

ni = Sample size in the strata a Ni = ni = nWi = n n n N Total attributes in the total sample divided by the sample size. For standard error of the estimate, we used the formula for simple random sample.

ˆ ˆˆ s.e( P ) = PQ / n in the case of proportion ) se( x ) = s / n where s is the estimate of standard deviation.

1.7.

Preparation for the implementation of the field survey

The preparation primarily involved pre-testing questionnaires and training the enumerators. Three types of pre-coded questionnaires were used:

6

(a) The street children questionnaire (Annex 3 Bengali version and English version Annex 4) (b) The guidelines for the key informants (Annex 5) (c) The employers questionnaire (Annex 6) The street children questionnaire was developed by ILO/IPEC and it was field-tested by FREPD. Based on the results of the field-testing, the instrument was further revised in a joint meeting of FREPD and BBS with technical approval from ILO/IPEC. The questionnaires are provided in the annexes.
Training Manual for the Data Collectors A training manual was prepared in order to facilitate the training of enumerators. The manual covered objectives of the study, concepts and definitions, selection of street children, explanation of some terms considered difficult by the team, methods to be followed in the data collection etc. Particular attention was paid to being sensitive to street children. 1.8. Field survey

The field survey started on 29.6.2003 and was completed on 28.7.2003. The study covered 450 street children.
1.9. Quality control mechanism and reliability of estimates

Quality control in data collection and management of data has been considered seriously. The following measures were adopted for ensuring quality of data. • Quality of Questionnaire: Care has been given to make the Bengali version of the questionnaire simple, clear, easily understandable and free from ambiguities. These were pretested before finalization. The test and pretest of the same questionnaire by two different data collectors were used to determine the reliability of the methods and investigators. • Data Collectors selected are masters degree graduates having sufficient experience in collecting data from children. Data Collectors were men. Each team had a supervisor, to supervise the field work, and edit the field-in questionnaires. Female enumerators could not be involved due to the fact the survey had to be conducted at night. Therefore, from cultural, safety and security point of view it was not feasible to engage female enumerators. All enumerators underwent rigorous training on conceptual and technical issues.
1.10. Field realities and limitations of the data

The sampling frame has been based on the estimates obtained in the QCS. There were some changes as regards the number of locations in a town and the number of street children living in a location between the QCS/enumeration (in Nov/Dec 2002) and the baseline survey (in July 2003). In some sample locations the required sample of street children were not found because of the mobile nature of their living and working conditions. In such situations the team moved to another neighbouring locations to interview the remaining sample units. The survey was administrated after 9 PM. After sun set, particularly after 8 PM sex workers (both male and female), tea and coffee sellers, children involved in drug business and the like remain busy either with customer or searching for the customer. Majority of such street children refused the study team to give any interview even after repeated requests. Consequently, 7

the sample has relatively lower representation of those street children involved in earning activities at night. It is to be noted that the above-mentioned street children generally do not sleep at night, but they sleep during day time at parks. Street children in general were found reluctant to give any interview because they were, at the time of interview, tired and sleepy. Moreover they fear being interviewed. Most street children, particularly younger children do not know their own exact age, the education level, occupation status, income level, and land holdings of their parents. They also could not remember when they first left the house. As a result, the validity of their responses and therefore of the data should be read with caution. A good number of street children, as reported by key informants, were involved in theft, snatching, sex work, drug business, pick pocket (Annex 7). During interview these children did not report these activities, instead reported other activities. This is true that such children have two occupations, one which they want to show to public so that no one could suspect them in their involvement in anti social activities. This category of street children also do not like to stay permanently in one locality. They often move to other locality, even to other town. The above field realities should be kept in mind while the reader interprets the results.

8

Chapter – 2 National Estimates of Total Street Children and Their Background Characteristics
The Baseline survey was administered during June-July 2003. The sample size (450 = n) was allocated to 17 regions using proportional allocation scheme. The percentage distribution of 2,573 (= N) street children by age, sex, location, etc. was based on the distribution of sample street children.
2.1. Estimate of street children by sex and by division

Out of 2573 street children 97.56 percent were boys and 2.44 percent were girls. Table 5 gives the distribution of street children by sex and division. Girl street children were found in 4 divisions. The survey did not find any street children in the Rajshahi and Khulna division. In Dhaka and Chittagong division about 3 percent were girl street children. The estimates of QCS, in our opinion, are underestimates as sex workers, drug abusers, pick pockets, theft, snatchers were under represented. The extent of underestimation could not be found out. The statistics of Table 2 shows that 54.8 percent of all street children alone live in Dhaka division, 14 percent each in Chittagong and Sylhet divisions. The percentage share of Rajshahi, Khulna and Barisal was small. Table 2 also gives the number of locations by division. As regards the number of location Chittagong ranks first while Dhaka ranks second, although Dhaka division had the highest percentage of street children.
2.2. Estimate of street children by regions

Bangladesh was previously divided into 23 districts. These districts are now known as greater districts, as each of them are further divided into more districts and currently there are 64 districts. Each of these greater district is termed as region by BBS in this survey (Table 3). Table 2 represents the distribution of 2573 street children by division and it can be seen that as Dhaka is the capital of Bangladesh and is the biggest city, it has the highest incidence accounting for 54.8 percent of the street children. Sylhet division and Chittagong account for 14 percent of street children population.
2.3. Home regions of street children

Table 3 gives the distribution of street children by the regions where they were living at the time of the QCS. The survey reveals that the poorest children migrated to the metropolitan cities of Dhaka, Sylhet and Chittagong from the other regions. The percentage distribution of street children by their original home district is given in Table 6a.

9

2.4.

Demographic characteristics of street children

The survey interviewed only those street children who reported their age between 5 and 17 years. The age distribution of street children is given in Table 5a. It can be seen that (48.7 percent) were of age 11-14 years, More than 26 percent of children were between 14-17 years old and 19 percent had age 8-11 years. Sex-wise variation was small. The mean and median age for boys and girls were as follows. Girls were relatively younger compared to boys. The three distributions are negatively skewed.
Sex Mean 12.19 11.49 12.16 Median 12.73 12.12 12.71 Mode 12.74 12.52 12.73

Boys Girls Total
Religion of street children

Among 2573 street children 98.7 percent were Muslims and 1.3 percent were Hindus (Table 9).
Parent's occupation

The distribution of street children by parent’s occupation is given in Table 10. The distribution by father's occupation identifies three main occupations. These are: (a) Nonagricultural labor, (b) Agricultural labor and (c) Small business. More than 70 percent street children’s fathers are labors either agricultural or non-agricultural (Table 10). As regards mother's occupation 62 percent mothers are housewife. It is important to observe that more than 21 percent mothers work as maidservant. The distribution demonstrates that the street children come from poorest families.
Education of parents

Table 11 provides the distribution of parents by their level of education. Nearly 70 percent fathers and 76 percent mothers were illiterate, 82 percent fathers and 88 percent mothers have no formal education. Comparing the national level of adult illiteracy rate of 40 percent, it may be inferred that street children come from families with lowest educational attainment.
2.5. Personal information about the street children

Table 12 provides the distribution of the reasons cited by street children for leaving home. The 5 most important reasons mentioned were: ♦ ♦ ♦ ♦ ♦ ♦ Poverty/hunger Ran-way from home Stepmother/father Earn money (income) No one to look after them Abuse : 38 percent children left home on account of this : 14 percent children reported this : 11 percent children reported this : 11 percent children reported this : 9 percent children reported this : 6 percent children reported this

These statistics reveal a gloomy picture. Most children left home because they were living either in abject poverty or in abject misery. They were forced to leave home when living with parents/relatives became intolerable. This may be termed as forced expulsion from the home or 10

push migration as used in demography. A small number (11 percent) reported pull factor (to earn money). It is to be mentioned here that about 13 percent children were orphans with both parents deceased. The survey reveals that 13 percent of all street children were orphan of both parents. Table 13 gives the estimated number of children who reported to the parents before leaving home for the town. The survey reveals that among the street children having at most one parent 50.4 percent did not inform while 49.6 percent informed the parents. Further, the survey found that 53 percent of children (Table 14) came to the town without the consent of parents. Out of this 53 percent, 47.3 percent came alone while 5.7 percent came with friends. It is also important to observe that although nearly 50 percent informed the parents not all of them took the consent of parents while coming to the town.
2.6. Duration of stay in the town

In terms of duration of stay in the city where they were being sampled, Table 15 shows that 16 percent were staying in the town for less than 3 months, nearly one third for less than a year, nearly 52 percent for less than 2 years and 48 percent for more than 2 years. As high as 23.4 percent had been living in the town for more than 5 years. A little less than 2 percent did not respond to the question. The average duration estimated is 34.6 months. The standard error of the estimate is calculated at 1.54 months.
2.7. Visit to the parents

The survey found that 57.4 percent of the children did not visit their parents (Table 16). Those who did not visit is comprised of two categories: One who had both parents dead and the other who had at least one parent alive. The percent of street children having both parents dead and did not visit was 13.1 percent. However, as high as 1095 (42.5 percent) visited parents. Among 1095 street children, 50 percent visited in last month, 18 percent in the last 3 months, 10 percent during last 6 months, 7 percent during last 12 months and more than 24 percent a year before. If visited and not visited are considered as indicators of keeping link with parents, then we see that nearly 58 percent have very weak link with parents and 42 percent children have been keeping links with parents.
Reasons for not visiting parents

The survey explored the reasons for not visiting the parents and these are presented in Table 17. The responses reveal that 24.5 percent did not visit because their parents were dead, 47 percent did not visit because they had no attraction to visit parents, 16 percent because they could not afford travel cost and 4 percent because their parents did not want their children to visit them. A close review of data of Table 17 reveals that 67 percent children and parents have detached themselves from one another.
2.8. Educational and training issues

The survey revealed that 82 percent of street children who had first job prior to street children stopped their school education in order to start work in their first job.

11

2.9.

Reading and writing skill

The street child was asked whether he (she) can read and write. If the answer was yes for reading then the child was asked to read a part of the questionnaire as a means of verification that the child can actually read. Thus, if the child could show its ability to read, then the child was categorized as able to read, other wise categorized as unable to read. Likewise the ability to write was determined through verification. The outcomes of the two tests are collated in Table 18a and Table 18b. The survey reveals that 76.4 percent street children could not read and only 23.6 percent street children could read. The level of skill was found to vary across division and by age. The highest percent (26.9) was observed for Rajshahi Division and also for 11-14 years age group children. The lowest (12.5%) level was observed for the Sylhet division and for the age group 511 years (12.0%). The survey reveals that only 24.2 percent could write and 75.8 percent could not write. The standard error of these estimates was 0.02. Writing skill was also found to vary over the division and age groups. The highest percentage (44.2%) was observed for Rajshai division and for 11-14 years age group (28.3%), and lowest was for Sylhet (14.3%) and 5-11 years age group (13%). Tables 18a and 18b show that high correlation understandably exists between reading and writing skills.
2.10. Schooling status of street children (ever attended school)

It was found that 60.7 percent children never attended any formal and non-formal school (Table 19). The standard error of the estimate was 2.30 percent. As high as 39.3 percent of the street children surveyed attended at least one type of school. Among all street children 29.7 percent attended formal school (27 percent exclusively formal schools) and 12.3% non-formal schools. In earlier chapter, it was observed that more than 82 percent of children who had attended school at some point stopped school even before they were street children. Table 19a shows that only 8 percent children were attending schools at the time of the survey. These statistics show that about 10 percent children stopped school during first job and the survey period.
2.11. Education

Table 19 reveals that 39.3 percent street children ever attended schools. The class completed by these 39.3 percent was different. Table 19b shows that 22.6 percent of school attending children just attended class I, but not completed class I, 28.8 percent completed class I, 23.2 percent completed class II, 13.0 percent completed class III, 6.7 percent class IV and 5.7 percent completed at least class V. The average years of education (completed) comes at 0.672 with standard error of 0.058. This implies that the overwhelming majority who joined street children had virtually no education.

12

2.11.1. Reasons for not attending schools in the past

About 60.7 percent street children never attended any formal or non-formal schools. The frequency distribution of the reasons they cited are given in Table 19c. Among the reasons the following four were found to be important: * * * * Parents did not send Parents/ Family could not afford Had to work for the family School education was not important : 44 percent of never attended children mentioned; : 31 percent of never attended children mentioned; : 9 percent mentioned; : 8 percent mentioned.

It is seen that poverty and ignorance about the importance of education were the main causes for not going to school.
2.11.2. Current status of schooling and class attendance

91.6 percent of street children were not attending schools and only 8.4 percent were attending schools (Table 19a). Among the school- attending children, 68.4 percent had been attending class I or pre-primary, 21.1 percent attending class II and 10.5 percent attending class III (Table 19d). These street children, few in number, combine school and work together. They mostly attend NGO run formal schools or Govt. and NGO supported non-formal schools.
2.11.3 Reasons for not attending schools currently

The respondents reported the reasons for currently not going to schools (Table 19e). The most important reasons were: * * * * Cannot aford school cost Because of work School education is not important Parents did not send, and now the school age is over : : : : 41 percent mentioned 38 percent mentioned 9 percent mentioned 7 percent mentioned

It can be inferred from the above responses that school for the majority is possible through motivational work and financial support.
2.12. Perception about importance of education

The respondents were found to have ambivalent ideas about the importance of education. However, they considered that it helps people to acquire life skill. Among all respondents 80.2 (Table 20a) percent reported that they give importance, while 19.8 percent reported that education and schooling "is not so important". However, when asked, would you go to school if you are given an opportunity, as high as 83 percent readily agreed that they would definitely go to school (Table 20b).

13

2.13.

Skill training and job preference

Street children are unskilled workers and the survey (Table 21) reveals that 95 percent of street children would avail of skill training if provided the opportunity and only 5 percent would not avail. In terms of what was their preferred work, most of the respondents said that the most preferred work is driver. The second preferred work is mechanics; and third preferred job is garage worker.

14

Chapter – 3 Past Work History of Street Children
3.1. First job of the street children

33.8 percent of the surveyed street children's first job is their current job. The remaining 66.2 percent street children had at least one job prior to their present job. Table 23 gives an account of their first job. ♦ 17.8 percent reported domestic servant as their first job ♦ 13.8 percent reported agricultural work as being their first job ♦ 34.6 percent reported jobs like begging, cooli, pickers of papers, hawkers, flower sellers, workshop workers, helpers etc. as being their first job
3.2. Age of the children when first started working

Nearly 42 percent of the street children started working even when they had not completed their 7 years. About 50 percent started working when they were in the age interval 811 years. About 8 percent started when they were of age between 11-14 years (Table 24). The mean age of starting the first job was 7.81 years and the standard error is 0.052 year. The mean age of street children was 12.16 years at the time of the survey. This indicates that on an average one street child had been working 4.35 years prior to the date of interview.
3.3. Reasons for starting the first job

A total of 1704 children's (66.2 %) first job is other than their current job. They reported the reasons for first starting work (Table 25). The most outstanding reasons reported by 88 percent children was poverty. About 5 percent children reported abuse by stepfather or stepmother.
3.4. Who put them to work in the first job

In terms of who put them to work in the first job, Table 26 shows that in case of 55 percent of the children it was their parents who put them to work in their first job. About 26 percent of the children reported that they themselves started the first job. In case of 20 percent of the children the relatives, friends, cousins, brothers and sisters put them to work in the first job. Recall that a significant number of children reported that their first work was being a domestic servant or an agricultural laborer. It can be inferred then that being a domestic servant puts the child at risk and makes them vulnerable to being on the streets.

15

3.5.

Whether the children stopped school to work in their first job

Another important finding of the survey is that 82 percent stopped going to school in order to start work in their first job (Table 27). The remaining 18 percent children did not stop school. They were attending school. They combined work and school together. From our experience this combination does not last. Ultimately they leave school for good and join the labor market on a permanent basis. In case of street children, the result was the same, all are now full time street children. Some however continue their education in non-formal schools.

16

Chapter – 4 Current Work, Work Conditions and Related Issues
4.1. Current work

The street children in the study areas are engaged in a wide variety of activities (Table 28a). Out of 2573 street children, most frequently mentioned activities are: collect old papers and other things (19.6%), cooli /minti tokai (14.7%) and begging (10.7%). Some activities with small frequencies have been grouped as others where 19.1% of the street children are engaged. Cooli are those who work at the railway station or launch ghat to assist the customers with their luggage while minti are those who work in the market places to assist the customers with their shopping. These two categories have been merged together because of small frequencies. The aforementioned activities were reported by the street children at the time of interview. The street children, particularly who are involved in activities not approved by the society suppressed the nature of their works. They did not report correctly. The key informants mentioned some activities, which were not mentioned by street children. These activities are: Theft, Pick Pocketing, Snatching, Sex work, Drug Business, Informer of anti social people etc. Table 28a presents percentage distribution of street children by their current work and geographical division. Cooli/minti was reported as the most significant activity in three divisions namely, Barisal, Khulna and Sylhet while in Chittagong, Rajshahi and Dhaka collection of old papers and other things came as the major category. Table 28b presents the percent distribution of street children by age and current work. In response to the question, how the street children got into the particular type of work, 75.8% responded that they themselves found the work, followed by 15.3% who reported getting the assistance of friends. There were some other minor sources, which are stated in Table 29.
4.2. Age when started working as street children

The age distribution in Table 30 shows that the modal age group for entry into the work force was 8-11 years. The mean age of entry was estimated at 9.6 years. The standard error of the etimate has been worked out to be 0.14 years.
4.3. Duration of street based work

The length of work of the street children varied widely from less than three months to more than five years (Table 31). Two-third of the children have been working more than one year, (26.7%) of them have been working between 2-5 years in the current job, and 21.2% have been working between 1-2 years. The average duration comes at 28.3 months. The standard error of the estimate is 1.23 months.
4.4. Daily working hours

Percentage distribution of street children by their daily working hours by division is presented in Table 32a. 37.6% percent reported to have worked between 8-12 hours a day whereas 32.7 percent reported working more than 12 hours a day. Division wise more than 50% of the street children of Khulna and Sylhet divisions reported to have worked more than 12 hours. 17

In Rajshahi division 46.2% reported to have worked more than 12 hours. In Barisal, Dhaka and Chittagong divisions majority reported to have worked between 8-12 hours. The estimated average number of hours worked is 10 while the mode of working hours is 10.5. The standard error of the estimated average hours is 0.16 hours. Table 32b presents percentage distribution of street children by their daily working hours and age. In the age group 5-