The study aims to analyze trends in the number of leading Angolan insurance companies for benchmarking purposes in the sector between 2020 and 2023, using the DEA-VRS window model. The analysis was conducted by calculating the efficiency scores of each insurer in two-year-wide moving windows, followed by a comparison of efficiency distributions over time and the identification of the insurers that served as benchmarks in each window for benchmarking purposes. The analysis was conducted by calculating the efficiency scores of each insurer in two-year movable windows, followed by a comparison of efficiency distributions over time and the identification of the insurers that served as benchmarks in each window for benchmarking purposes. The results reveal a trend of increasing average efficiency scores and a concomitant reduction in the number of insurers identified as benchmarks, a pattern that was consistently evident across the three analysis windows. In practical terms, the findings also indicate that the universe of benchmarks is becoming more concentrated, which requires attention to the periodic updating of the analysis windows and the selection of variables to ensure that benchmarking continues to reflect efficient practices within the analyzed sample. The article stands out for exploring a little-studied market, using robust primary data, offering practical value by identifying benchmark insurers, and revealing an unexpected inverse relationship between average efficiency and the number of benchmarks DMUs.

This study investigates the dynamics of relative technical efficiency across Angolan universities from 2017 to 2022, aiming to establish operational benchmarks and support strategic decision-making in higher education. Data were collected from eighteen officially recognized public and private institutions using annual statistical reports published by MESCTI. The methodology relies on non-parametric intertemporal models, specifically DEA-VRS and Window DEA, which capture performance variations across consecutive time windows. Results include efficiency scores per institution and time frame, input and output slacks required to achieve technical efficiency, and identification of reference peers for benchmarking purposes. Findings reveal a consistently declining trend in average efficiency over the period, with UCC being the most frequently referenced unit. Critical variables such as enrollment and course offerings underscore the need for diversification and alignment with labor market demands. Estimated slacks offer actionable insights for policy and cost optimization. As a methodological extension, the study proposes investigating the relationship between variable-level inefficiencies (SBM) and global efficiency scores (CRS) to better understand interwindow performance trajectories.