Principal investigator: Viviane Pires Ribeiro
Article title: Effects of FUNDEF/B on school attendance, school flow, and child labor: An analysis based on the 2000 and 2010 censuses.
Article authors: Gabriela Cruz and Rudi Rocha
Location of the intervention: Brazil
Sample size: 2.995.211 individuals
Main theme: Education
Type of Intervention: Impacts of FUNDEF/B on Brazilian basic education
Main Variable of Interest: Educational outcomes and child labor
Evaluation method: Experimental Evaluation (RCT)
Context of the Evaluation
The 1988 Constitution of the Federative Republic of Brazil established that 18% of the Union's revenue and 25% of the revenue of states and municipalities, derived from taxes and transfers, should be allocated to education. However, according to Cruz and Rocha (2018), these resources tended to be misapplied or even directed to other areas due to the lack of earmarking of these revenues for specific purposes and poor oversight. In light of this situation, the Federal Government introduced the Fund for the Maintenance and Development of Basic Education and the Enhancement of the Teaching Profession (FUNDEF) in 1998, with the aim of promoting greater equity in educational spending in Basic Education, thus altering the distribution of resources between states and municipalities.
Subsequently, in 2006, FUNDEF was replaced by the Fund for the Maintenance and Development of Basic Education and the Enhancement of Education Professionals (FUNDEB), which then included all of Basic Education – daycare and preschool, elementary school, high school, and Youth and Adult Education (EJA). This increased the proportion of taxes and transfers allocated to the fund and incorporated new sources of revenue and new minimum spending levels per capita for education. In this sense, Cruz and Rocha (2018) state that the introduction of FUNDEF/FUNDEB induced the expansion of educational spending in poorer municipalities, and can be considered one of the most important changes in the financing regime of basic education in Brazil.
Intervention Details
Cruz and Rocha (2018) evaluate the impacts of FUNDEF and FUNDEB on school attendance, age-grade distortion, and child labor among children exposed to these resources during school age. The research was conducted using two main data sources: FINBRA (Finanças do Brasil/STN), which contains self-reported information regarding the annual revenues and expenditures of Brazilian municipalities from 1993 to 2010; and the Demographic Censuses of 2000 and 2010. The total sample used in the study consists of 2.995.211 individuals, representing a population of 27.788.831 children aged 7 to 14, considering the years 2000 and 2010 together.
Initially, the authors constructed a database of municipalities over the years to document the relationship between the implementation of FUNDEF/FUNDEB and the educational expenditures of municipal governments. To this end, variables related to education spending and net FUNDEF/FUNDEB transfers were created. Subsequently, another database was constructed by cross-referencing information on children aged 7 to 14 with the finances of the municipalities where they lived during school age, in order to assess the impacts of education spending on individual outcomes. Since the microdata from the 2000 and 2010 Demographic Censuses contain information on outcome variables – school attendance, age-grade distortion, and child labor – and control variables, the authors constructed the variable "average net FUNDEF/FUNDEB per capita" for each individual, based on data on municipal finances.
Methodology Details
Cruz and Rocha (2018) estimate fixed-effects regressions for municipalities and cohorts based on microdata from the 2000 and 2010 Censuses, and on municipal FINBRA expenditures from 1993 to 2010. The authors explore the variation in the exposure of different individuals to education resources from FUNDEF and FUNDEB, which occurs between municipalities and birth cohorts. The hypothesis is that, conditional on the fixed effects of municipality, age, Census year, and controls, the remaining variation in the variable of interest is orthogonal to any variation in latent determinants of the outcome variables. Thus, models are estimated with balanced samples (according to municipalities over time) and unbalanced samples; and with a sample considering only children whose status in the household is "child" versus a total sample of children.
Results
The results found by Cruz and Rocha (2018) reveal positive effects of increased resources transferred via FUNDEF/FUNDEB. Considering municipalities where the proportion of enrollments in the municipal network is equal to the median (27% in 1995), an increase in net FUNDEF/FUNDEB transfers of R$100 per capita is related to an increase in the probability of a child attending school by 0,05 percentage points, while the average age-grade distortion decreases by 0,02 years. These results are equivalent to an impact of, respectively, 0,9% and 6,5% of a standard deviation compared to an increase in net transfers of one standard deviation (approximately R$385). The effect on the probability of a child working is close to zero. However, in municipalities where all enrollments belong to the municipal network, the effects are much more significant. An increase of R$ 100 in net transfers from FUNDEF/FUNDEB implies a 0,3 percentage point increase in the probability of attending school, and a reduction in the age-grade distortion of 0,09 years.
Regarding the differentiated impacts according to the education level of the head of household, the results reveal that the positive effects tend to be stronger for those from poorer family backgrounds. For them, an increase in the net per capita FUNDEF/FUNDEB of R$ 100 causes an increase in school attendance of 0,07 percentage points and a reduction in the learning gap of 0,03 years, considering the median proportion of municipal enrollments. That is, in addition to the total positive effects on attendance and flow indicators, FUNDEF/FUNDEB also had important impacts on reducing inequalities in the indicators among children.
Lessons in Public Policy
What are the impacts of basic education funding on educational outcomes and child labor? The results found by Cruz and Rocha (2018) pointed to positive impacts of increased resources transferred via FUNDEF/FUNDEB on school attendance and age-grade discrepancy. These effects are greater in municipalities where the proportion of enrollments in the municipal network was initially higher. On the other hand, the impacts on child labor are, in general, close to zero.
Regarding the differentiated impacts according to the educational level of the head of household, the results found by the authors show that children from lower socioeconomic status (heads of household without primary education) benefit most from the increase in resources allocated to education. Thus, it is likely that, in addition to the positive impact on attendance and flow indicators in general, the redistribution of resources via FUNDEF/FUNDEB also has an effect of equalizing children's outcomes, which means equalizing at least some of the opportunities offered to them. In this sense, Cruz and Rocha (2018) argue that this is an important aspect in terms of public policy formulation, considering that the transfer of resources via FUNDEF/FUNDEB mainly benefits children from families with lower socioeconomic status, who are the most vulnerable and have the worst educational outcomes.
References
CRUZ, Gabriela; ROCHA, Rudi. Effects of FUNDEF/B on school attendance, school flow and child labor: an analysis based on the 2000 and 2010 Censuses. Estudos Econômicos (São Paulo), v. 48, n. 1, p. 39-75, 2018.
