Different teachers for different students? Evidence on teacher-student matching and its consequences in the Italian case

We present novel evidence for Italy thanks to a unique dataset linking students standardized achievement measures and their teachers' characteristics, both for language and mathematics. First, we explore the extent to which teachers are allocated to students from different family backgrounds, focusing on observable teachers' characteristics that parents and school leaders could interpret as signals of their effectiveness. We test whether these characteristics affect achievement by means of student fixed-effects models that exploit the «within students-between subjects» variation in test scores. For each teacher, we compute a summative measure of Cumulative Contribution to Student Achievement (CCSA) and we study its allocation among students from different educational backgrounds. Results show that classes with students from a lower social background are more frequently staffed with teachers possessing less desirable traits, especially in secondary schools and for language teachers. These traits affect achievement and also teachers' CCSA is unequally allocated among students. Our analyses also show that teacher-students matching is stronger in Southern regions, although not for all grades and subjects. Evidence suggests that also in Italy, unequal allocation of students to teachers could be a hidden form of educational inequality.