Chapter 16However this is not the end of the story. For, as we have
seen, the overall happiness of the population is strongly af-
fected by how people behave to each other. So, when we
study individuals, we need to understand not only how their
own happiness is determined but also what determines
their behavior to others. We have few good measures of be-
havior. Whether people attract a partner is of course affected
by their behavior, but by much else besides. The nearest we
have to a measure of behavior is a person’s criminal record.
So in column (2) we analyze what determines the number
of times people have been convicted or cautioned by the
age of 30. The best predictors of this are their qualifications,
and, not surprisingly, their behavior as a child.
Does allowing for these effects alter the relative ranking
of the different dimensions of child development? From
other studies we know that on average each crime reduces
aggregate life- satisfaction in the population by 1 point- year.
So we could compute an aggregate effect of each dimension
of child development on human happiness by combining
column (1) aggregated (over say sixty years) with the nega-
tive of column (2). Aggregating a number over sixty years
makes it quite large, even if it is quite small on a per year
basis. Given that, the numbers in column (2) would not re-
verse our previous conclusion that emotional health is the
most important dimension of child development for pur-
poses of aggregate human happiness.^11
How Parents and Schools Form the Child
So what determines how children turn out to be by the age
of 16?^12 By far the best evidence on this comes from the Avon