Document Type


Publication Date



We estimate the impact of increases in schools constructed from 1985 to 1995 on girls educational outcomes in Nepal. We use a difference-in-differences framework by combining the across-district differences in the number of new schools with variation in exposure to these schools created by the virtue of individuals being of school-going-age during the school construction period. Our results indicate that the construction of an additional school (per 1,000 square kilometers) increased the probability to read and write among females by 1.5 percentage points and increased their highest level of schooling attained by 0.12 units but did not affect basic literacy skills among males. Our back-of-the-envelope calculations suggest that, on average, the increase in the number of schools can explain about a fourth of the total differences in the reading and writing outcomes between females who were of schooling age during the period of school constructions and those who were not. These results underscore the continued importance of increasing access to schooling in developing countries like Nepal.'