| Abstract: |
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. |