The success of private schooling in Britain in delivering high academic achievements and better jobs has recently been attributed to the schools engendering high aspirations and confidence among their students. However, no empirical evidence has yet been offered to support this diagnosis or the policy implications that may follow. Using recently repaired data on secondary school type from the 1970 British Cohort Study and on cognitive skills at age 16, we investigate the effect of private schools on pupils’ self-esteem and locus of control at ages 10 and 16, and on aspirations and high-value network access at age 16. We then examine the effect of these factors on earnings in mid-career and their hypothesised significance for understanding the private-school earnings premium achieved in later life.
We find that private school pupils have substantially higher levels of self esteem, a more internal locus of control, higher job quality aspirations, higher occupational aspirations, and better perceived access to high-valued networks for job search.
Despite this association, self esteem is not raised by private schooling, once social background and prior cognitive skills are controlled for. In contrast, there is some evidence that private schools do help to stimulate locus of control, especially at primary level, and to raise aspirations and provide access to networks. This conclusion assumes, however, that remaining unobserved factors affecting both the decision to use the private sector and children’s non-cognitive attributes result in only small biases in the coefficients.
We find that, while locus of control and aspirations both have modest effects on pay at age 42, neither self esteem nor networks are linked to pay in later life. Moreover, only a small part of the private school earnings premium is accounted for by all these non-cognitive factors. Much of the premium is due, rather, to educational attainments. After controlling extensively for post-16 academic attainments, there is no significant residual premium for women. There remains, however, an unexplained pay premium for men.
While research on other non-cognitive attributes and school type is warranted, the main policy implication of these findings is that strategies to strengthen self-esteem, locus of control, aspirations and social networks in the state sector will contribute little to the objective of greater social mobility. Policies should therefore remain focused on narrowing socio-economic gaps in educational achievement.