A comparison between the primary school support provided by positive and negative outliers municipal governments in Brazil

• In Event: Leveraging the middle-tier of education systems for teaching, learning, and equity in Latin America

Proposal

Less than half of 10-year-old students achieve minimum proficiency levels in Brazil. Despite that, because of its federative system, some states and municipalities have much higher average student outcomes than equivalent subnational governments with similar socioeconomic conditions. The behavior of these positive outliers may influence other middle-tier and school leaders to improve their students' results.

This research applied mixed methods to analyze the support provided by municipal governments to primary schools in two contexts: positive and negative outliers. First, I used a hierarchical linear model with data from 2015 to 2019 to analyze the influence of practices implemented by municipal governments to support primary schools on student learning achievement and, based on that, to categorize municipalities into three groups: positive outlier, average, and negative outlier. Then, I selected four positive outliers paired with four negative outliers from the same state to conduct semi-structured interviews with key stakeholders: the head of the municipal education department, municipal department staff, and school leaders.

The findings from the quantitative analysis show that most practices implemented by municipalities are statistically significant contributors to students’ learning. Establishing a focus on student achievement improvement is the practice with the largest contribution, followed by providing high-quality professional development to teachers and school leaders and using learning data as evidence for planning and organizational learning. The results are aligned with previous studies from high-income countries and robust for the conditions tested: mathematics and language test scores, among and within states, the percentage of students with adequate learning, and timewise variation in test scores. Using a threshold of two standard deviations, around 180 municipalities were categorized as positive and 200 as negative outliers out of 4927 in the sample.

As for the implementation analysis of the practices to support schools, three main insights were found. First, all four positive outliers implement all eight practices considered. This means that effective Brazilian municipalities similarly support their primary schools to those from high-income countries, with practices such as promoting instructional leadership and using learning data to plan interventions. Second, that implementation is limited when compared to those practices in the context of high-income countries. For each of the eight practices considered, positive outliers in Brazil lack the implementation of certain aspects described in the literature from high-income countries. For instance, professional development initiatives are not part of an explicit strategy, nor do they consider the needs of specific teachers. Third, there are consistent differences in the implementation of five of the practices when comparing positive with negative outliers, while three practices show inconsistent differences. This evidence supports the findings from the quantitative analysis, explaining the differences in several practices, such as establishing a shared vision to improve student achievement, which is mentioned more frequently by stakeholders from positive outliers. At the same time, practices such as promoting good communication and relationships were not significantly different for all the pairs of municipalities compared. This shows how the context may mediate the contribution of specific practices by middle-tier governments on improving student achievement.

Author

• Leite Dalmon Danilo, Kobe University