While aspirations are receiving increasing attention in the study of poverty, empirical evidence remains limited. Using Young Lives data for Andhra Pradesh, this paper investigates whether mothers' aspirations matter for their children's education outcomes. We observe a strong relationship: aspiring to one additional year of schooling lifts the grade achieved at age 15 by 1.8 years on average. The relationship is, however, non-linear, with low aspirations having low effects, medium aspirations having large effects, and high aspirations having slightly lower but still large effects. These impacts remain strong and significant after controlling for a wide range of village, household, parent and child characteristics, including cognitive skills, and after allowing for endogeneity using a control-function estimation with as identifying instruments whether the child is a first-born son, as well as the mother's assessment of the usefulness of education for herself. Aspirations tend to have smaller effects for children from wealthier backgrounds and higher educated mothers, and impact also depends on the village setting.

Extending the analysis, we find similar effects on teenage mathematics and verbal test results. Mothers' educational aspirations for their children have, however, no effect on the choice to attend a private (versus government) school, no effect on attending school at age 15, and a negative effect on working versus staying at home at age 15. Aspirations are higher for mothers from wealthier households and for mothers with higher education, while village effects also play a role. The findings shed light on recent theoretical work and provide new insights on the channel through which intergenerational mobility takes place, indicating that maternal aspirations, rather than maternal (or paternal) education, drive household investment in child education.