Using data from three 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 relevant reforms taking place at the national level over the last century and affecting the working conditions of primary school teachers, matching them by the year they entered the profession. After showing that teacher experience/age and qualification are significantly correlated with student competencies, we study the correlation between teacher working conditions (including recruitment, pay and retirement policies) and pupil achievement. Our identifying assumption is that the impact of reforms dissipates with the distance between the reform's introduction and entry into the profession. The results point to a more selective recruitment process and, to a lesser extent, more generous reward policies as effective ways to enhance student performance.