This study investigates effects of a conditional cash transfer programme (CCT) in Brazil – Bolsa Família (BF) – on school outcomes, particularly children's achievements on standardised tests, pass-grade rates, and dropout rates. The educational conditionality of the programme, requiring enrolment in school and minimum school attendance, figures as a major justification for public investment in BF. It is expected that BF will reduce short-term poverty and boost children's human capital, thus inducing long-term socioeconomic improvement. In order to achieve its long-term objective, BF should be able to improve not only enrolment and attendance rates, but also learning outcomes and grade promotion amongst beneficiary children. However, these effects, particularly learning outcomes, have not yet been reported in the literature. The hypothesis investigated in this thesis is that length of time of participation in the programme and per capita cash amounts received by families are key variables in assessing BF's effects on children's educational outcomes. As the programme improves household income, requires a high rate of school attendance, and monitors children's health and nutritional conditions, a positive effect on children's performance should be expected over time. Similarly, the amount of cash paid to families should have an impact on changes induced in the home environment that are beneficial for children's educational outcomes. Empirically, the dissertation combines three national datasets from governmental agencies for the years 2005 and 2007. These data contain information on test scores in Portuguese Language and Mathematics for fourth grade pupils, school context, and BF parameters (intake, time of participation, and cash value), which are used in cross-sectional and panel analyses to test the above hypotheses. The results show that although beneficiaries tend to attend less well-resourced schools, the influence of individual and household characteristics on test scores overshadow that of school resources, suggesting that demand-side interventions might result in gains in children's performances. The cross-sectional analysis at the school level suggests that BF's contribution to school outcomes depends on the length of time of participation and the per capita cash value paid to families. In addition, these two BF parameters have substitute effects, that is, as the per capita cash increases, school performance increases; however, the contribution of time of participation to gains in school performance diminishes and vice-versa. As a sensitive analysis to test the direct effects of length of time of participation and per capita cash on school outcomes, a subsample was used, which includes only schools in which more than 80% of pupils are beneficiaries. Results from this subsample confirm the positive effects of time and cash on school outcomes, although only cash is statistically significant. Furthermore, a school-and-time fixed effects model is estimated using panel data for 2005 and 2007 for the same school outcomes. The results also suggest that improvements in school outcomes are expected over time as a result of exposure to the programme, although this varies across regions. The findings support the idea that improvements in educational opportunities and outcomes for children of low-income families in Brazil require a non-educational policy measure – the reduction of the immediate income poverty – as intended by BF. Nevertheless, there is also an urgent need to address inequalities in standards of education supply and special attention should be given to children whose families are recipients of BF in promoting access to pre-school programmes. Even though educational policies are necessary, they are insufficient to promote human capital amongst the poorest families in Brazil. In this sense, CCTs do not represent an opportunity cost for educational policies. Instead, they are important allies in promoting education access and equity.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:558154 |
| Date | January 2012 |
| Creators | Simoes, Armando Amorim |
| Publisher | University of Sussex |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://sro.sussex.ac.uk/id/eprint/40673/ |
Page generated in 0.0019 seconds