Abstract:
This paper uses value-added models and panel data from a comprehensive set of highstakes secondary school exams to assess determinants of student performance during the coronavirus pandemic in Bhutan. Gender gaps, urban-rural gaps, and socioeconomic gaps do not appear to have widened substantially after the pandemic-related closure. Student characteristics (gender, access to a computer at home) and school characteristics (boarding facilities, urban location, class size, computers) predict performance. Quantile regression analysis shows that home learning environment (parental education) is a predictor of Class XII performance for higher-performing Class X students. Computer ownership at home clearly improves English performance but the pattern is less clear for Dzongkha, which has limited software and internet content for instruction. Influence of past performance and gender is stronger for students in the lowperforming group. Boarding facilities played a unique role in Bhutan's pandemic response, potentially mitigating learning losses and offsetting household differences that condition education outcomes.