Principal investigator: Pedro Jorge Holanda Alves
Article titlePublic spending on education and school performance.
Author of the articleJoana Monteiro
Location of the Intervention: Brazil
Sample Size: Stage 1 (4.115 observations) and Stage 2 (220-240 observations)
Sector: Education
Type of InterventionThe effect of public spending on school performance.
Main Variable of InterestTuition fees and academic performance
Evaluation method: Experimental Evaluation (RCT)
Policy Problem
It is possible to verify that in Brazil, school inclusion policies have been positive in relation to increasing the number of enrollments and student retention. According to Monteiro (2015), in 20 years, the average years of schooling of the Brazilian population increased by 55%, so that 93% of children between 6 and 14 years old are enrolled in primary education. However, according to the results of the Programme for International Student Assessment (PISA), out of the 65 countries that took the exam, Brazil ranked 58th, 55th and 59th in the rankings of the highest scores in mathematics, reading and science, respectively.
Similar to the positive results achieved by public policies that incentivize school enrollment, programs focused on student performance also generate benefits for students. To achieve this, policymakers must adopt more measures aimed at improving the performance of enrolled students and, in a way, generate more efficient results. In certain situations, it is necessary to allocate greater spending to education in order to implement these measures.
In Brazil, the budget allocation system dictates that mayors have autonomy to define their spending, but revenue collection is the responsibility of the federal government, which collects a large portion of the revenue and transfers proportional shares to states and municipalities. In some specific cases, it is possible to identify municipalities with similar characteristics that receive federal resources in different amounts.
Assessment Context
Certain municipalities in coastal regions benefit from being located in areas rich in oil royalties, and for this reason, these municipalities receive a certain amount of transfer from the federal government. Monteiro (2015) analyzes the impact of public spending on the educational quality of Brazilian oil-producing municipalities that have benefited from an increase in royalty revenues. Educational quality can be positive in two ways: on the one hand, spending can generate an increase in the number of enrollments and ensure that a greater number of children are enrolled in schools, and on the other hand, spending can improve student performance.
Due to the sharp increase in revenue received from royalties in 1998, the analysis begins in that year (some variables were used in 2000 as it is the closest year with available data) and continues until 2010, the last year with information from the Demographic Census. Variables related to school enrollment are represented by the net enrollment rate in primary education (the ratio between the number of people aged 6 to 14 who are regularly enrolled in primary education and the total population in the same age group) and the percentage of children aged 6 to 14 who do not attend school. To measure the progress in the schooling of the younger generation, the author uses the illiteracy rate of individuals aged 11 to 14, the percentage of children aged 6 to 14 with two or more years of school delay, and the expected years of schooling. Finally, the analysis will seek evidence to determine if the receipt of oil royalties is associated with these improved educational outcomes, using data from the Prova Brasil (Brazil Test) as an indicator of school performance.
Policy Details
Due to increased oil production, the Brazilian government established the Petroleum Law, which stipulated that municipalities would benefit from a portion of these extractions and that part of these oil royalty payments would not be exclusive to the producing municipality, but also to nearby regions. The weight of these transfers became significant after 1997, with production more than doubling between 1997 and 2010, increasing the transfer from R$424 million to R$21,6 billion during that period. It is expected that these increased transfers will generate more revenue for municipalities, consequently creating opportunities for the adoption of policies focused on student performance.
Methodology Details
For his estimates, Monteiro (2015) separates the evaluation process into two stages. Initially, using indicators of school quantity and quality in a simple regression, he investigated the performance of municipalities that invested the most in education between 2000 and 2010. In the second stage, he verified the existence of a relationship between municipal revenue gains from oil production and increased spending on education. The affirmation of this relationship allows the use of regression with instrumental variables to identify which additional gains due to oil activities have impacted investment in education and school outcomes.
Based on geographical criteria, the Petroleum Law applies the distribution of royalty revenues, with municipalities located on the coast entitled to receive more than neighboring municipalities inland. This distinction allows for the definition that municipalities on the Brazilian coast are "treated" because they receive a larger amount of revenue than their neighboring municipalities, which would be defined as "control" municipalities. The use of the law becomes favorable for causal analysis, since geographically they are similar municipalities, with the difference being that there are distinct revenue gains in relation to oil royalties.
Although this strategy is not without its critics, since the determination implies that receiving or not receiving this resource is independent of the economic and political characteristics of the municipalities, only in a maritime border that has oil production. In this way, municipal policymakers cannot define the level of royalty revenues each year, and the difference between the two groups of municipalities would be defined by the increase in expenditure caused by the increase in revenues transferred by the federal government.
However, since we are interested in seeing the effects reflected in expenditure, even though 75% of the resources received are legally mandated to be invested in education, the decision on how to allocate them depends on the efforts and objectives of the mayors, which can introduce some bias into the estimates. For this reason, in this second part, a simple estimation was not made, but rather the value of oil production was used as a tool to capture the local impact of education expenditure.
Results
In the first stage, encompassing approximately 4.000 municipalities, the results indicate that increased spending on education leads to a higher number of children in schools, using enrollment rates and out-of-school indicators. This result likely stems from the constitutional provision of the Fund for the Maintenance and Development of Basic Education (FUNDEB), which links federal government transfers to municipalities based on enrollment numbers. In the second stage, however, estimates show that due to increased budgets from royalties, municipalities raise the average teacher salary by 9% in an attempt to improve student performance; however, there is no indication that this policy has resulted in improvements in educational performance.
Although a portion of oil revenues has been allocated to education spending, evidence from 239 observations between 1998 and 2010 suggests that the sector has not been highly prioritized, with only 14% of royalty revenues going to education and a difference of only 13% in municipal spending in the Brazilian coastal region compared to its neighbors. In monetary terms, the increase in the value of oil production is associated with a growth in royalty revenue of R$0,02, which consequently resulted in municipal spending and generated an increase in education spending of R$0,003.
Lessons in Public Policy
In some cases in Brazil and around the world, we encounter situations where local governments, for some external reason, receive a larger amount of financial resources and, therefore, have greater availability of investment and spending options. These cases serve as experiments to show how local governments with a large amount of revenue, or with unexpected revenue gains, behave when deciding how to allocate their expenses.
This work contributes to the literature by showing how Brazilian municipalities perform in terms of educational quality when they have greater availability of resources. The results presented show that the lesson to be learned indicates that simply using resources for higher teacher salaries does not yield significant results for performance. It is necessary for municipalities to adopt other policies, such as better teacher training, greater student support, or improvements to school infrastructure.
Reference:
MONTEIRO, Joana. Public spending on education and school performance. Brazilian Journal of Economics, v. 69, n. 4, p. 467-488, 2015.
