Résumé
This paper estimates the causal effects of math instructional time on student achievement using within-student variation across math topics in the TIMSS 2011 data. Based on the assessment's detailed measurement of math skills, it shows that each weekly hour of math instruction in a given topic increases student test scores in that topic by 4.3% of a SD. However, the productivity of instructional time varies significantly based on the implementation of teaching practices that prioritize student active participation: teachers emphasizing these practices are more than twice as productive as other teachers. This result holds true regardless of student gender, social origin, and is observed internationally in 42 countries.