[8] Many of Lavy's publications are co-authored with Joshua Angrist and typically exploit natural experiments to estimate economic outcomes such as the returns to schooling.
[9] Throughout the mid-1990s, Lavy performed field research on health and education in several developing countries, including Morocco, Côte d'Ivoire, Pakistan, and Ghana.
Analysing how the quality of healthcare in Ghana affects the health of rural children, he (together with John Strauss, Duncan Thomas and Philippe de Vreyer) finds that making birth services and other related child programmes more widely available and improving water and sanitation infrastructure would likely strongly reduce the mortality rates of children in rural areas.
In his most-cited study, Lavy (with Joshua Angrist) uses the application of Maimonides' rule in Jewish schools to estimate the effect of class size on student achievement in a quasi-regression discontinuity design and finds that class size reductions significantly and substantially increase the student achievements of fourth and fifth graders, though not of third graders.
[17] Another substantial body of Lavy's work on education economics concerns the effect of financial incentives for teachers and students on learning outcomes.
[23] Together with Joshua Angrist and Analia Schlosser, Lavy exploits widespread parental preferences for mixed sibling compositions and ethnic differences therein in Israel to assess the relationship between household fertility and children's education; they don't find any evidence for a trade-off between quantity and quality.
With regard to gender peer effects in Israeli elementary, middle and high schools, Lavy and Analia Schlosser find that increasing the proportion of girls in a class improves male and female pupils' cognitive outcomes by decreasing classroom disruption and violence, improving inter-student and student-teacher relationships, and decreasing teachers' fatigue, though the change in composition in itself doesn't affect individual behaviour.
[26] This result is largely replicated in Israel, where Lavy, Schlosser and Daniele Paserman study the effect of the proportion of low achievers in a class on their peers and find again a negative impact of bad peers on their classmates, especially on classmates who are low achievers themselves, mostly by disrupting teachers' teaching, deteriorating inter-student and student-teacher relationships, and making violence and classroom disruptions more likely to happen.