Displacement, deprivation and hard work among Syrian refugee children in Lebanon

Background The protracted Syrian war resulted in the largest refugee crisis of our time. The most vulnerable are children who face separation from parents, interruption of schooling and child labour. This study explores the living and working conditions of Syrian children in Lebanon. Methods In this cross-sectional study, we randomly selected 153 informal tented settlements and conducted interviewer-administered surveys among Syrian refugee working children in the Bekaa Valley in Lebanon. Those aged 8–18 completed a questionnaire on sociodemographic and occupational characteristics; those aged 4–8 years were surveyed through a household questionnaire. Results We surveyed 1902 households, including 12 708 individuals and 4377 working children. Female-headed households were poorer and more food-insecure than male-headed households. Among working children (4–18 years), the average age of starting work was 10.9 years and 74.8% worked in agriculture. Compared with boys, girls earned less and were less likely to be enrolled in school. For 96.3% of working children aged 8–18 years, forced exodus to Lebanon was associated with a first child labour experience. Working conditions were harsh and worse for girls who compared to boys were less likely to receive their salary on time and take time off work. Girls worked longer in the sun and cold and were more likely to report having a health symptom at work, working under pressure and using sharp or heavy objects at work. Seventy-nine children reported knowing another child who died following a work accident. Conclusion Children, as young as 4, are forced to work, and many are compelled to forgo educational opportunities in favour of harsh and harmful labour due to difficult economic conditions. State policies facilitating access to work for adult refugees will help families meet basic needs and decrease their dependence on child labour as a coping strategy.

The presence of each item was given a score of one, so the maximum score was 17. We categorised the infrastructure problems index into three problems or less, four to six problems, and more than six problems. [1][2][3] Assets A household assets index was calculated based on several questions about ownership of 27 items (goods and commodities) such as: bed, washing sink, cooking stove, refrigerator, heater, television, radio, car, telephone line, and others. A principal components analysis was conducted where indicators of household assets were converted into z-scores and factor loadings for a single assets factor were calculated. For each respondent, values of the indicator variables were multiplied by the factor loadings to obtain an assets score. Respondents were grouped into quartiles by assets score, with the lowest quartile considered the poorest.

Food security
Food security was firstly measured through the Reduced Coping Strategies Index (RCSI), which includes five individual coping behaviours (eating less preferred foods, borrowing food/money from friends and relatives, limiting portions at mealtime, limiting adult food intake, and reducing the number of meals per day). The RCSI is used to measure the presence and severity of food security across different contexts and settings. 4 Food security was also measured through an adapted version of the Livelihood Coping Strategy Index 5 using ten long-term coping strategies:  Selling household goods  Spending some or all household savings  Buying food on credit or borrowing money to purchase food  Borrowing money  Selling productive assets or means of transport  Reducing essential non-food expenditures  Withdrawing children from school  Selling house or land  Asking for money from strangers  Accepting high risk or exploitative temporary jobs

Health symptoms at work
Health symptoms experienced at work were listed in the survey including: fatigue, fever, body weakness, imbalance, tachycardia, cramps, physical breakdown, quivers/chills, chest pain, yellow/pale skin, dehydration, cold sweat, difficulty in concentration/memory loss, nose bleed, skin allergies, respiratory problems, stomach problems, urinary problems, and others. Each symptom was transformed into a binary variable (No: 0, Yes: 1). The sum of all 19 symptoms was computed for each child, then the mean number of symptoms was computed by sex.

Work injuries
Working children responding "Yes" to any of the following injuries were considered to have a work injury. Responses were structured as a four point Likert scale (very satisfied, satisfied, not so satisfied, and not satisfied at all). Respondents answering "Not so satisfied" or "Not satisfied at all" were given a score of one; the remainder were given a score of zero. A higher score was indicative of having less satisfaction with life.

Child wellbeing
A Child Wellbeing Index was developed from 16 statements that asked about wellbeing:  I feel alone these days Responses were structured as a four point Likert scale (strongly disagree, disagree, agree, strongly agree). Respondents answering "Agree" or "Strongly Agree" were given a score of one; the remainder were given a score of zero. A higher score was indicative of worse child wellbeing.

Child perception
A Child Perception Index was developed from six statements that asked about perceptions of the future:  I will learn the craft I like  I will receive the degree I like  I will find a job I have a degree for  I will become a successful person  I will remain jobless for a long time  My future will become better Responses were structured as either "Agree", "Disagree", or "I don't know". There were minimal respondents who answered "I don't know", so respondents who answered negatively about their perception of the future were given a score of one for each statement. A higher score was indicative of a more negative perception of the future.

Child optimism
A Child Optimism Index was development from 10 statements adapted from the Life Orientation Test-Revised. 6  In uncertain times, I usually expect the best  It's easy for me to relax  If something can go wrong for me, it will  I'm always optimistic about my future  I enjoy my friends a lot  It's important for me to keep busy  I hardly ever expect things to go my way  I don't get upset too easily  I rarely count on good things happening to me  Overall, I expect more good things to happen to me than bad Responses were structured as either "Agree", "Disagree", or "I don't know". There were minimal respondents who answered "I don't know", so respondents who answered negatively about their feelings were given a score of one for each statement. A higher score was indicative of a more negative feelings.

Quality control
Following recruitment and survey completion, each household was re-visited by a Quality Control Team to check whether (i) the tent was actually visited by the data collector, (ii) the female homemaker was interviewed whenever possible or was replaced by another member of the household, (iii) every working child (aged >8 to ≤ 18 years old) in the household was interviewed, and (iv) whether the month and year of birth for family members were reported according to legal papers.
Although our target sample size was 1,884 households, we sampled 1,907 eligible households of which four refused participation and one was vacant, leaving 1,902 in the final sample. The Quality Control Team detected and corrected inconsistencies in 8% (n=152) of the households, mainly relating to the year of birth of household members. *Give information separately for exposed and unexposed groups.