全球智库信息集成服务系统

• Those implementing services must have a clear incentive in the programme either through the provision of direct support to them under the programme or through being held to account.
• The capacity of the implementing agency is not directly associated with the quality of implementation. In some cases, small may be better.
• Deeper integration between Bolsa Familia and health and education services is associated with better reported implementation.
• More disenfranchised (poorer and living in remote or rural geographies) populations are generally enrolled successfully in Bolsa Familia but generally receive worse education and health services.

Bolsa Familia (BF) is a conditional cash transfer programme introduced in 2003 in Brazil. This report aims to understand the principal agent problem in BF. We do this by running three generalised linear models to find associations between contextual factors such as capacity of agents against outcome measures related to the quality of the provision of services such as the quality of registries. Our findings suggest that those implementing services must have a clear incentive in the programme either through the provision of support under the programme or through the accountability process. Smaller municipalities and those with poorer constituencies tend to have more accurate registries. Health and education services seem to have less incentive to operate in BF as they have no electoral incentive or receive support under BF. Finally, our models show that more integration between BF and health and education services is associated with better reported implementation.

• Chapter One

  Introduction

• Chapter Two

  Data

• Chapter Three

  Approach

• Chapter Four

  Findings

• Chapter Five

  Limitations

• Chapter Six

  Conclusion