Reducing gender gaps in education remains a significant policy concern in many developing countries. One strategy that has been advocated by a range of international organisations to improve learning outcomes for girls is to increase the representation of female teachers. This has been supported by recent empirical evidence from India (Rawal and Kingdon 2010; Muralidharan and Sheth 2015). This extended essay examines whether the finding that female teachers improve learning outcomes for girls holds in other cultural contexts, drawing on longitudinal data on primary school students from Vietnam and Pakistan.

Using dynamic OLS value-added models, I find that teacher gender has no significant impact on learning outcomes in Vietnam while in Pakistan male teachers significantly improve overall test scores in private schools for both boys and girls by around 0.13 standard deviations per year. This latter finding does not appear to be driven by positive sorting or differences in teacher characteristics or effort levels between male and female teachers. However, male teachers did not have a significant impact on overall test scores in public schools suggesting that the results may be driven by differences in unobservable characteristics between male and female teachers in private schools.

Finally, in both public and private schools in Pakistan, male teachers had a stronger impact on test scores in Maths than English and Urdu, although no such effects were found for Vietnam. This finding suggests that so-called ‘stereotype effects’ may play an important role in teacher-student gender interactions in some developing countries.