Abstract
The quality of primary school education is often still low in developing countries. The reading intervention EGRA Plus in Liberia is an attempt to improve teaching through teacher training and increased accountability. We show that the effects, that a previous evaluation found to be large, are considerably heterogeneous and mostly benefit the best students. Moreover, we show how the intervention changed the behavior of students, parents, teachers and principals in order to assess the causal channel. We find no evidence of positive or negative spillovers to other subjects like mathematics.