Assessing the pure technical efficiency of public primary schools in the City of Zagreb

Abstract

This paper is the first to assess the pure technical efficiency (PTE) of 111 public primary schools in the City of Zagreb during the 2022/2023 school year, using an input-oriented Data Envelopment Analysis (DEA) – BCC model. PTE was measured using two input variables (expenditures and number of teachers) and three output variables (average grades, number of pupils who passed the class, and secondary school enrolment points). The analysis revealed substantial variation in efficiency scores, with 20% of schools identified as fully efficient. The average efficiency score was 0.90, suggesting that, on average, schools could reduce inputs by 10% while maintaining current output levels. A follow-up super-efficiency DEA model was applied to rank efficient schools and identify best practice examples that could serve as models for others. Based on super-efficiency scores, schools were grouped into five performance categories – best practice, efficient, near-efficient, emerging, and low efficiency. Additional findings suggest that larger schools tend to be more efficient and that benchmarking within peer-size groups can help uncover high-performing schools that might otherwise remain under the radar. The results underscore the importance of context-sensitive efficiency evaluation in education and highlight the need for improved data on school characteristics to support more comprehensive, actionable
assessments.