Child protection risks child poverty and social transfers

Child protection risks, child poverty and social transfers Paul Dornan March 2013

Outline 1. Young Lives study 2. Extent of child protection related risk in survey data and links with material circumstances 3. Tracing impact of poverty/material circumstance on child protection concerns 4. Implications – children, protection and social transfers

Young Lives study Cohort study, pro-poor, collecting data on children in Ethiopia, India, Peru and Vietnam • Around 3, 000 children in each country. Information collected at child, household and community level • Core survey instrument so far conducted at 2, 5, 8 years; and 8, 12, 15 years • Qualitative and quantitative (household and school) data Multidimensional approach and wide range of data collected. At child level this includes information on physical health, cognition and wellbeing. Survey data is publically archived. Strands of ongoing research include: extent and effects of inequalities on children; life course analysis (what matters and when); how the wider context shapes child development

Extent of child protection related risk in survey data and links with material circumstances: child work 7 6 Most work unpaid, caring and on family farm/business. Variations by gender, wealth level and urban/rural 5 4 3 2 1 0 Least Poorest Chronic Female poor (wealth index) (poorest index) age 12 and 15) Male Urban Rural Activities outside the household for pay Tasks on family farm or business Doing domestic tasks Caring for others Ethiopia, young people aged 15. Hours per day

Extent of child protection related risk in survey data and links with material circumstances: stunting % of young people stunted 50 45 40 35 30 25 20 15 10 5 0 Least poor (age 15) Poorest (age 15) Chronic poor (poorest age 12 and 15) Urban Poorer children and those in rural areas most likely to be stunted Rural Ethiopia, young people aged 15

Extent of child protection related risk in survey data and links with material circumstances: subjective wellbeing % reporting I am proud of my shoes or of having shoes 80 70 60 50 40 30 20 10 0 Poor children report lower levels of subjective wellbeing Least Poorest Chronic Female poor (age 15) (poorest age 12 and 15) Male Urban Rural Ethiopia, young people aged 15

Extent of child protection related risk in survey data and links with material circumstances: experience of physical violence 100 90 80 70 60 50 40 30 20 10 0 % reporting I have never been hurt physically Least Poorest Chronic Female poor (age 15) (poorest age 12 and 15) Male Urban Poorer children do not report more experience of violence (though may be underreporting) Rural Ethiopia, young people aged 15 Self Administered Questionnaire

Tracing impacts: poverty and child protection concerns Work injury. Poor children face between 1. 8 times (Ethiopia) and 2. 9 times (Vietnam) greater chance of reporting being injured at work than non-poor children (age 15) Food shortage. Children in households that reporting an episode of food shortage: scored lower in cognitive achievement tests in Andhra Pradesh and Ethiopia; and positioned themselves lower on the subjective wellbeing ladder in Ethiopia and Peru (at age 15) Malnutrition. Stunting at age 8 predicts lower school aspirations, selfefficacy and self-esteem at 12 years old (eg. controlling for other factors, stunted children reported expectation of leaving school earlier than non-stunted children (by about 2. 4 months) (pooled data from 4 countries)) Environmental shocks. Depending on the type of shocks and where it happens and sibling composition, either increase or reduce hours of child work (eg. in Andhra Pradesh income shock increases hours of work for girls by 3. 3 hours per week)

Conclusions # 1: child poverty and child protection • Child protection risks are widely reported by Young Lives children (both quantitative and qualitative data). Such concerns are interlinked/multidimensional (for example, links between low income, ill health and family strain) • Variations in risks/ opportunities experienced by children systematically associated with household background. Associations are not always straightforward, and (likely) reporting problems with survey data • Earlier poverty-related risk linked with negative effects on key child protection concerns, therefore reasonable to expect improved material circumstances pay off in improvements in some child protection indicators

Conclusions #2: child protection and social transfers • • Since poverty presents protection concerns, social protection has important potential to be (broadly) pro-child protection. Plus social floor/ systems approach = social policy development. Beyond that, a core question of how social protection pays off for children (in context) – design implications for transfers? • Employment guarantees and labour displacement – child care facilities; use of direct transfers; • Categories very useful – why so few child benefit examples? But care also – YL evidence in Ethiopia suggests OVC in Ethiopia less of a predictor of education and health outcomes than poverty would be; • Balance in use of conditionality, quite a lot may be achieved by cash transfer alone, linking services and so on. Care conditionality doesn’t shape unhelpful stereotypes; • Dignity and scheme design (eg. how children feel about receipt, stigma and so on, likely to be connected with labelling, breadth of coverage etc).

Haymanot’s story, poverty, ill-health, work and early marriage • Haymanot was age 12, she was living with her aunt in rural Ethiopia and attending school in a nearby town. When her single mother suffered successive bouts of illness, Haymanot returned home to rural Tigray to care for her • Haymanot’s mother could no longer work, the family could no longer pay daily labourers and turned instead to sharecropping. This meant that they only received half the produce, resulting in food shortages • Initially, Haymanot attended school in the afternoon and worked on the PSNP in the morning with her sister. However, Haymanot’s sister also became sick. A combination of famine, drought and sickness ran down the household assets, and Haymanot dropped out of school in grade five. She described her sadness as she was winning awards but says it was her decision to stay and care for her mother • Haymanot was also suffering from repeated malaria, diarrhoea, vomiting and fever, exacerbated by her work in stone crushing factory. Three years later, age 15 Haymanot married. She feels that ‘it has benefited me because I have got rest from going to work. ’ She said her health was better and ‘we have enough farm products’ so is not suffering from food shortage as when she was with her mother.

Targeting by category (OVC, Ethiopia) • In Young Lives households one in five children had lost one (usually fathers) or both parents by age 15 • Inequalities in schooling and health outcomes greater between children living in urban and rural areas and according to household consumption level than between orphans and non-orphans • But varies by age of child (when orphaned): younger orphans (single and double) more likely to be enrolled than non-orphans but older orphans less likely to enrolled than peers

Transfers significant, but ongoing coverage gaps 80 70 60 50 40 30 20 10 0 Least poor Poorest Chronic poor (poorest age 12 and 15) % of children receiving any transfer urban rural % Government transfer Ethiopia, households of 15 year olds