Selecting or rewarding teachers? International evidence from primary schools

Using data from the existing four waves of PIRLS, this paper examines the effect of teacher quality on fourth-grade students’ literacy test scores by exploiting variations induced by reforms in teachers’ selection and/or reward schemes. We construct an original data set of reforms by the government taking place at the national level after World War II and affecting the working conditions of primary school teachers. We match these indicators of reform activities with the year of entry into the profession for each teacher, our identifying assumption being that new entrants are mainly affected by the reform, with the impact of the reforms dissipating over time. We show that more selective recruitment processes are positively correlated with pupil achievements, while better teacher working conditions (including to some extent retirement policies) exhibit the opposite correlation.