Filipe Recch, Postdoctoral, Stanford University
Mid-level managerial practices and student performance gains – learning from Brazil’s decentralized education state systems
The main research question behind this project is: to what extent and how do mid-level management practices in regional offices of education relate to gains in student achievement in Brazilian states? Initially based on the state of São Paulo, and now expanding to other states in Brazil - currently, Ceará, Pernambuco and Minas Gerias - the idea is to take advantage of the decentralized structure of state educational systems in Brazil in order to isolate differences in managerial practices carried on in different regional offices that operate under the same policies, mandated by each state. First, the value added by each state regional offices of education is estimated in order to explain, at least partially, the differences in achievement within the state system. Then, based on these estimates for various student cohorts, consistently high- and low- performing regional offices are identified and the leadership in each of these offices are interviewed using the World Management Survey (WMS). Finally, based both on the WMS scores and qualitative data collected during the interviews, the main differences in managerial practices between both groups are identified. The goal is to establish which managerial practices tend to have a positive impact on student performance.
I will present the findings for the state of São Paulo and initial results considering Ceará and Pernambuco. The findings for São Paulo show that state regional offices do influence student achievement and do so differently for state and municipal schools. The evidence gathered using the World Management Survey (WMS) show that two managerial areas are markedly different between high- and low-performing regional offices: targeting, meaning the decision making related to what schools to provide more support to, and role assignment in relation to personnel organization within the teams in the regional offices. In both cases, high performing offices tend to present better scores and more autonomous decision making related to these areas. Additionally, the qualitative analysis shows that high-performing offices had better understanding of the pedagogical situation of their schools and act according to these diagnostics, while low-performing ones are excessively dependent on the central office for support and diagnostics. The results of Ceará need revision and Pernambuco's are still in very early stages.
My goal with this presentation is to discuss not only the findings we have so far, but how to strengthen the analysis and how to use the differences between states from an analytical perspective.