Chapter 14The Effects of Individual Secondary Schools
Our aim is to explain child outcomes at age 16: emotional
(as measured by reports from parent and child, averaged);
behavioral (as measured by the parent); and intellectual
(the GCSE score). We control for the child’s relevant out-
come at age 11 (before they entered the secondary school),^2
as well as for family background. We then estimate the im-
pact of each school by entering a separate dummy for every
school. Table 14.1 shows the standard deviation of these
school effects.
The effect of the school is remarkable— not only on in-
tellectual performance, as Hanushek found, but also on
happiness (emotional health) and behavior. For example,
we can take a child of given happiness at age 11 (and given
family background) and ask how much happier did that
child become at age 16 if in a school at the 83rd percen-
tile of happiness production compared with a school at the
50th percentile of happiness production. The answer, as
Table 14.1 shows, is that the child became happier by a mas-
sive 0.26 standard deviations of happiness.^3 In other words a
child who at 11 was at the median level of happiness would
by 16 have been at the 60th percentile. This is a remark-
able impact. The effect on behavior is nearly as great, but
here we want to stress the effect on happiness— since many
Table 14.1. How children’s outcomes at 16 are affected by secondary
school attended (ALSPAC) (β- coefficients)Emotional Behavioral Intellectual
Secondary school 0.26 (.01) 0.21 (.01) 0.29 (.01)