‘Graduation programmes’ are anti-poverty interventions that aim to set its participants on a virtuous cycle out of poverty. They provide a comprehensive package of support that often includes consumption transfers, asset transfers, access to savings and credit, training and coaching. Training and coaching are mostly focused on income generating activities but also include messaging regarding health, sanitation and nutrition. Existing research suggests that programmes have positive effects on household living standards and assets, including consumption, food security and asset holdings. However, the impact on individual household members, including children, is far less understood.

This report presents findings from a quantitative evaluation assessing the impact of the Chemen Lavi Miyò (CLM) – pathway to better life – programme on child development and wellbeing in Haiti. The CLM programme is implemented by Fonkoze and constitutes a ‘graduation programme’. Over a period of 18 months, it provides intense material and non-material support to women who have dependents – often young children – and are living in extreme poverty in rural Haiti.

The evaluation is based on a quasi-experimental design, with a control group having been sampled after the treatment group had already been established. Propensity Score Matching (PSM) is used to generate a counterfactual that allows for difference-in-difference estimates. Key outcome indicators of interest represent main risk factors for child development. These include household living standards (as proxy for poverty), biological risk factors(such as nutrition and health) and psychosocial risk factors (such as maternal depression, early learning and safety and security). Analysis also explores the balance between paid work and unpaid care and the potential tension that may arise between providing quality care while engaging in income generating activities at the same time.

The CLM programme has a positive effect in various areas of child development and wellbeing, namely household living standards, access to food, maternal mental health, education, exposure to harsh corporal punishment and violent scenes at home, and occurrence of children in the community being sent to work as domestic servants. Effects are largest in outcome areas that are strongly related to availability of economic resources, such as perceptions of whether there is enough money to provide for children and access to food. Large impacts are also observed for indicators that are strongly related to support that is directly provided through the programme, such quality of roofing and drinking water.

The programme has limited impact in relation to other areas that are important for child development. Child feeding practices, undertaking stimulating activities with children, and attitudes regarding child disciplining are not affected by the programme. Certain areas of positive engagement with children – such as asking them about school, work and friends – are positively affected, while others – such as praising children – are not. Given the set of messages that are provided throughout the programme period, it is surprising to see no impact on child feeding practices. Lack of impact in areas of stimulating activities and child disciplining is more plausible as the programme does not focus on these areas.

Finally, the programme contributes to changes in women’s time use towards spending more hours on paid work away from care work. Children appear to fill the gap as they increase their engagement in care work over the course over the programme period. This is a highly gendered and age-based shift: it involves mostly daughters and mostly young children aged 5-9. Children who are involved in care work do not tend to spend more time on these activities, however. This suggests that the main concern may not be those children engaged in care work, but rather those that they care for.