Understanding the Quality of Implementation in Bolsa Familia

This paper aims to promote a greater understanding of how the 'black box' of implementation in a conditional cash transfer (CCT) programme interacts with the quality of implementation. The 'black box' of implementation refers here to programme and external factors that could impact the outcome of the programme such as: the capacity of municipalities; the supply of services; the integration of services; geography; political motivations; levels of poverty; and urban versus rural context. By taking data from federal Brazilian datasets related to these factors and administrative data on the quality of implementation from the Bolsa Familia (BF) programme, this paper models associations between the key variables. The findings suggest some nuanced findings on capacity and the nature of poverty in municipalities, which seems to suggest that overall resources are perhaps less important than the approach taken in a municipality to BF implementation. The supply-side appears to matter, both in terms of the quality of services associated with beneficiaries meeting BF conditionalities as well as how they are provided. As a result, the paper suggests that those designing CCTs should perhaps invest more in the supply of services. Bolsa Familia in Brazil and similar CCT programmes have been copied in Latin America and other continents and countries (e. g. the Philippines, Indonesia, Uganda, India). The evidence on CCTs increasingly concludes that they are successful in raising household expenditure and promote the take-up of health and education services, contributing in some cases to significant reductions in poverty rates. Impact evaluations have increased the understanding of the relationship of the coverage of beneficiaries, targeting of households and incentive levels (level of benefits) in a CCT programme with the effectiveness of these programmes at household level. 1 However, the understanding of which programme and contextual factors influence the success of these programmes remains limited. This in turn limits knowledge on what works in what context, which is of crucial importance as CCT programmes spread across the globe. This paper seeks to shed further light into the 'black box 'of a CCT by reviewing relevant Brazilian federal datasets and modelling associations between key variables. The 'black box' of implementation refers to a range of programme and external factors that could impact the outcome of a CCT programme. Programme variables can refer to the capacity and resources of those implementing BF as well as the provision of health and education services that beneficiaries engage with. External variables consist of political, demographic and geographic factors. With regards to outcomes, the paper looks at the municipal level of implementation and examines the quality of decision-making at the local level in registering beneficiaries and monitoring conditionalities. In short, the paper wants to understand better how the decentralised approach used in Brazil is contributing to programme outcomes.