In international comparison, German university graduates are elder on average when they enter the labor market. Apart from the longer duration time of study at university, one fundamental reason is the long secondary schooling time. Therefore, almost all of the German federal states have introduced a reform of shortening the length of secondary schooling by one year while holding the curriculum almost constant. Hence, the reform has increased the learning intensity ratio, i.e. the ratio of academic curriculum content per unit of instructional time, for the treated students considerably. The educational policy reform was at first enacted in 2003 and realized in 2007 with a double cohort of graduates in the German state of Saxony-Anhalt. We use this reform as a natural experiment to evaluate the causal effects of higher learning intensity ratios on student performance. Identical final written exams for both cohorts allow us to assess school performance directly. We use achievement grades at graduation in mathematics, German literature and English to approximate the effects on human capital accumulation. The effects of increasing learning intensity ratios on student performance depend on subjects and differ by gender. We find significant negative effects on student performance in mathematics for females and more pronounced negative effects for males. Student performance in foreign language has also decreased due to the reform, but only for females. In contrast to that, no differences are obtained on grades in German literature for both genders. Our findings suggest inefficient learning intensity ratios in the acquisition of linguistic skills. However, increasing the academic content considerably in a predefined period of instructional time in mathematics cannot compensate for shortening schooling time. Moreover, some students are not able to cope with the increase in learning intensity. Perhaps, lowering the learning intensity ratio in such demanding subjects like mathematics by additional instructional time at the expense of subjects with low learning intensity is a reasonable recommendation. In addition, a revision of the curricula could be an appropriate response. Our results suggest that institutional features, such as learning intensity, matter. Policy makers should turn their attention from raising the quantity of education to raising the quality.


Büttner, Bettina
Thomsen, Stephan Lothar


student performance, school duration, learning intensity, natural experiment