This paper sheds light on a potential risk of trade-off between access to primary education and the quality of education in underserved areas in the case of Yemen, which still faces challenges in access, equity, and quality in spite of its rapid improvement over the past decade. It will first examine the level of learning achievement in relation to an improvement of enrollments and gender parity in rural schools where girls were more disadvantaged in 2004, when Yemen started the national basic education strategy toward 2015. It will also examine the level of learning achievement in relation to the other supply-side variables, of which quality and quantity are often affected by expansion of access. The data used here are the student- and school-level data collected by the JICA Research Institute at rural schools in Yemen in 2011 and the data from the TIMSS in 2011. The analyses point to a trade-off between improved access and quality of learning. The results indicated that the school-level increase of gender parity in enrollment and the total enrollment growth of boys and girls alike over the past three years are negatively associated with the current students’ math test scores. The estimation results of trade-offs are valid while controlling for basic students and family characteristics, such as parental education and occupation. As anticipated from previous education production function analyses, some policy variables, such as teacher availability and existence of participatory school management committees, change across schools, and they are important for explaining the learning differences in the relatively underserved rural areas of Yemen. These findings underscore the need of revisiting such policy measures in further increasing access while ensuring the quality standards for disadvantaged areas and avoiding widening of the quality difference within the country and ensuring learning for all.
Keywords: quality of learning, universal primary education (UPE), gender, class size,