The drivers of month of birth differences in children's cognitive and noncognitive skills : a regression discontinuity analysis

This paper uses data from a rich UK birth cohort to estimate the differences in cognitive and non-cognitive skills between children born at the start and end of the academic year. It builds on the previous literature on this topic in England by using a more robust regression discontinuity design and is also able to provide new insight into the drivers of the differences in outcomes between children born in different months that we observe. Specifically, we compare differences in tests that are affected by all three of the potential drivers (age at test, age of starting school and relative age) with differences in tests sat at the same age (which are therefore not affected by the age at test effect) as a way of separately identifying the age at test effect. We find that age at test is the most important factor driving the difference between the oldest and youngest children in an academic cohort; highlighting that children born at the end of the academic year are at a disadvantage primarily because they are almost a year younger than those born at the start of the academic year when they take national achievement tests. An appropriate policy response in this case is to appropriately age-adjust these tests. However, we also find evidence that a child's view of their own scholastic competence differs significantly between those born at the start and end of the academic year, even when eliminating the age at test effect. This means that other policy responses may be required to correct for differences in outcomes amongst children born in different months, but not necessarily so: it may be that children's view of their scholastic competence would change in response to the introduction of appropriately age-adjusted tests, for example as a result of positive reinforcement.