Search

With the rise of polarization and extremism, the question of how best to transmit civic virtues across generations is more acute than ever. In this paper, we test the hypothesis that schools can be the place for this transmission by empowering students and gathering them around concrete and democratically chosen objectives. We draw on an RCT implemented in a large sample of middle schools in three European countries. The evaluated program leads students to carry out collective citizenship projects in their immediate communities under the supervision of teachers trained in student-centered teaching methods. The program significantly increases student altruism, their political self-efficacy as well as the quality of their relationship with their classmates and their respect for the rules of school life (less sanctions and absenteeism). In all three countries, the benefits are greater for students with the highest level of altruism and interest in politics at baseline. Investments made at an early age appear to be complement to those made during adolescence for the production of civic virtues.

The Human Capital Theory developed by Gary Becker in the 60's substantially widened the area of investigation of economics. Over the last two decades, many studies in the economics of education intended to identify the characteristics of an educational system which enable individuals to acquire as much skills, knowledge and information as possible. This thesis contribute to this literature by studying two aspects of the educational environment that has particularly attracted economists' attention over recent years: teacher productivity and peer effects in the classroom. The first chapter of this thesis investigate to what extent teaching practices implemented by math teachers in the US relate to their students' math performance. First, it shows that every single hour spent in the classroom studying mathematics generate a significant improvement in students' math performance. Second, it shows that the productivity of instructional time strongly relates to the implementation of interactive teaching practices, which require student active participation in the lesson. More precisely, each hour spent with a teacher putting a high weight on this kind of practices is 2 to 3 times more productive than an hour spent with a teacher putting a higher weight on traditional practices, such a teacher lecture. The second chapter of this thesis studies the impact of a public policy aimed at improving teachers' practices, namely the individual teacher evaluation system in French secondary education. In this chapter, we show that students' performance at the end-of-middle school national exam significantly improve after the evaluation of their math teacher, not only for students taught by an evaluated teacher the year of the evaluation, but also for students taught by the same teacher on subsequent years, suggesting a long-lasting improvement in teacher pedagogical skills. These positive effects persist over time for students, who not only perform better at the end-of-middle school exam but also choose more often and graduate more often from the science track in high school. In addition, the positive effects of teacher evaluation are particularly salient in education priority schools, in contexts where teaching is often very challenging.The third chapter of this thesis investigates the effect of school peers' gender on students' performance and educational careers in French middle schools. First, it shows that the proportion of female peers' in middle school has persistent effects on students' educational careers as it not only affects students' test score at the end-of-9th-grade national examination, but also influences their track choices and high school graduation rates several years later. Second, it shows that a larger share of girls in the classroom has positive effects for girls and negative effects for boys. More specifically, it reduces girls' dropout rates and increases their probability to graduate from an academic track in high school, especially in the scientific track, while it increases boys' probability to attend a vocational school after 9th grade and decreases their high school graduation rate.; Initiée dans les années 1960 par l’économiste américain Gary Becker, le développement de la théorie du capital humain a considérablement ouvert le champ d’investigation de la science économique. Au cours des deux dernières décennies, de nombreux travaux de recherche en économie de l'éducation ont cherché à identifier les caractéristiques du système ou de l’environnement scolaire qui permettent aux individus d’acquérir un maximum de compétences, de savoirs et d'informations au cours de leur scolarité. Cette thèse s’inscrit dans ce courant de recherche et étudie plus spécifiquement deux aspects de l’environnement scolaire qui ont retenu l’attention des économistes : la productivité des enseignants et l’influence des camarades de classe. Le premier chapitre de cette thèse étudie dans quelle mesure les pratiques pédagogiques des enseignants aux États-Unis permettent d’expliquer les différences de performances de leurs élèves en mathématiques. Dans un premier temps, nous montrons que chaque heure passée en classe à étudier les mathématiques engendre une progression significative des élèves dans cette discipline. Nous montrons ensuite que la productivité de l’heure d’enseignement est très fortement corrélée avec la mise en place de pratiques pédagogiques interactives, qui requièrent une participation active de la part des élèves. Plus spécifiquement, chaque heure passée avec un enseignant mettant l'accent sur ce type de pratiques est 2 à 3 fois plus productive qu'une heure passée avec un enseignant mettant l'accent sur des pratiques plus traditionnelles telles que le cours magistral.Le deuxième chapitre de cette thèse étudie l’impact d’une politique publique visant à améliorer les pratiques des enseignants, à savoir le système d’inspection individuelle des enseignants du second degré en France. Dans ce chapitre, nous montrons que les performances des élèves en maths au Diplôme National du Brevet (DNB) s’améliorent significativement à la suite d’une inspection de leur enseignant de mathématiques, non seulement pour les élèves assignés à l’enseignant l’année de l’inspection, mais également pour les élèves assignés à cet enseignant les années suivantes, suggérant une amélioration durable de ses compétences pédagogiques. De surcroît, l’inspection des enseignants de maths de 3ème produit des effets bénéfiques persistants chez les élèves, qui se traduisent par une augmentation de leur probabilité de choisir une filière scientifique en première et d’obtenir un baccalauréat scientifique au cours des années suivantes. Finalement, les effets bénéfiques d’une inspection sur les performances des élèves en maths et sur leur trajectoire scolaire sont particulièrement marqués pour les enseignants de l’éducation prioritaire, lesquels font face à des contextes d’enseignement plus difficiles.Le troisième et dernier chapitre de cette thèse étudie les effets du genre des camarades de classe de 3ème sur le parcours scolaire des élèves en France. Deux ensembles de résultats se dégagent de l’analyse. D’une part, l’influence des camarades de classes est persistante au cours du temps, puisque la proportion de filles parmi les camarades de classes en 3ème influence non seulement la réussite au brevet, mais également le taux de décrochage scolaire, le choix des filières après le collège et le taux d’obtention du baccalauréat. D’autre part, la proportion de filles parmi les camarades de classes en 3ème a des effets bénéfiques sur la scolarité des filles alors qu’elle a des effets négatifs sur celle des garçons. Plus spécifiquement, cette proportion réduit le taux de décrochage scolaire des filles après la 3ème et augmente leur taux d’obtention d’un baccalauréat général, particulièrement dans la filière scientifique. A l’inverse, elle augmente la proportion de garçons choisissant une filière technique après le collège et réduit leur taux d’obtention d’un baccalauréat général.

In France, secondary school teachers are evaluated every six or seven years by senior experts of the Ministry of education. These external evaluations mostly involve the supervision of one class session and a debriefing interview, but have nonetheless a direct impact on teachers' career advancement. In this paper, we show that these evaluations contribute to improving students' performance, especially in math. This effect is seen not only for students taught by teachers the year of their evaluations but also for students taught by the same teachers the subsequent years, suggesting that evaluations improve teachers' core pedagogical skills. These positive effects persist over time and are particularly salient in education priority schools, in contexts where teaching is often very challenging. Overall, a system of light touch evaluations appears to be much more cost-effective than more popular alternatives, such as class size reduction.

In France, secondary school teachers are evaluated every six or seven years by senior experts of the Ministry of education. These external evaluations mostly involve the supervision of one class session and a debrie_ng interview, but have nonetheless a direct impact on teachers' career advancement. In this paper, we show that these evaluations contribute to improving students' performance, especially in math. This e_ect is seen not only for students taught by teachers the year of their evaluations but also for students taught by the same teachers the subsequent years, suggesting that evaluations improve teachers' core pedagogical skills. These positive e_ects persist over time and are particularly salient in education priority schools, in contexts where teaching is often very challenging.

Though teachers are consistently found to play a major role in determining student achievement, little is known about what teachers can do to increase their instructional productivity. This paper develops a new empirical strategy, based on within-student within math variations in student test scores, to assess the instructional hourly productivity of math teachers in the US. Building on these estimates, we show that teachers’ hourly productivity strongly relates to the use of teaching practices emphasizing student active participation in the lesson (modern practices). One weekly hour of math instructional time increases student test scores by 4.4% of a standard deviation on average, but one hour spent with a teacher above the modern practices index median is more than twice as productive as one hour spent with a teacher under this median (+5.9% vs +2.7% standard deviations). A further investigation suggests that the positive effects associated to modern practices are partially mediated by an improvement in student self-confidence and motivation to learn mathematics.