U.S., and Vietnam, respectively. We examine these effects on reading and math performance, as the literature reports varying findings for both subjects. While the effect of temperature on math performance has been found to be negative in most studies (e.g., Cho, 2017; Garg et al., 2020; Graff Zivin et al., 2018; Park et al., 2020), the literature offers contradictory findings for reading performance. With the exception of Park et al. (2020), who find similar results for reading and math performance, several studies report negative effects of temperature on math performance but not on reading performance (e.g., Cho, 2017; Garg et al., 2020; Graff Zivin et al., 2018). Last, we contribute to the literature by exploring heterogeneity in the effect of temperature on student test scores regarding the construction year of the school building. The literature finds heterogeneous effects of temperature on educational performance among students regarding gender (e.g., Chang & Kajackaite, 2019; Cho, 2017; Cook & Heyes, 2020; Dang et al., 2024), race and income (e.g., Park et al., 2020), and age (e.g., Dang et al., 2024; Zhang et al., 2024). Heterogeneity in the construction year of the school building, however, has not been previously explored in the literature. We argue that this construction year of the school might serve as a proxy for the overall quality and condition of the school building.1 Compared to newer school buildings, older school building are presumably less likely to have air conditioning systems and score lower on energy efficiency and insulation quality. Indeed, we document a positive relationship between the school building construction year and the energy label of the school. Although we 1Palacios et al. (2022) explore the impact of the physical state of schools on primary school students’ learning outcomes using field data from a large-scale study. Additionally, Palacios et al. (2020) examine the broader benefits of healthy buildings on productivity. 113
RkJQdWJsaXNoZXIy MjY0ODMw