出版社：Suntory Toyota International Centres for Economics and Related Disciplines
摘要：In order to assess the roles of schooling and educational qualifications in the emergence of adult social exclusion, a series of detailed regression models were explored separately for men and women for each of a wide range of indicators of adult disadvantage at both ages 23 and 33, including experience of unemployment, being in receipt of non-universal benefits, low income, low occupational class, living in social housing, and a high malaise score, as well as ever being a lone mother; a number of other measures were considered at age 23, such as experience of not being in education or employment for 16 to 23, early parenthood, or at age 33, such as lacking a telephone or cigarette smoking. The core strategy for assessing these issues was to consider successive blocks of characteristics, ordered by their increasing proximity to the level of qualifications and the adult outcomes. Thus, we began by considering a cluster of variables that represent the parental background of the survey members, including measures of childhood poverty, of father’s social class, parental housing tenure, parental education, and experience of family disruption. The second block of variables included reports of mother’s and father’s interest in their child education, filtered through the observational lens of a teacher, and on the child’s behaviourattributes, observed through a parental lens. Once selected terms had been added to the initial model for parental background, the resultant model was then used as an anchor for considering a third wave of potential characteristics for inclusion, which included summaries of educational test scores and of frequent school absences at ages 7, 11, and 16, and of three different reports of contact with the police by age 16. These terms were then also locked into the next model and reported qualification levels were then considered for inclusion. Two dominant patterns emerge from the examination of the wide range of outcomes considered. ! Firstly, educational qualifications show a clear and strong relationship to every single adult measure of disadvantage at ages 23 and 33 and both for men and women, which is generally stronger at age 33 than at age 23. This strong relationship emerges net of controls for a wide range of childhood factors, which includes measures of parental interest in education, results of educational test scores, and indications of frequent school absences. ! Secondly, the childhood precursor that most frequently remains a clear predictor of negative adult outcomes, net of all the other factors considered is childhood poverty. A number of other important findings emerge: ! Mother’s interest in schooling is more salient for women, whilst father’s interest matters more for men; ! Low parental interest in schooling, frequent absence from school, and low educational test scores are all quite influential on subsequent disadvantage, even net of qualification levels; ! Early contact with the police is a better indicator of anti-social behaviour than frequent absences from school in relation to adult outcomes for men, but absences are more influential for women. Specific continuities in exclusion also emerge: ! The father being in Social Classes IV or V remains a clear predictor of male survey members also being in these Classes at ages 23 and 33, net of all the other factors considered; ! Growing up in social housing shows a similar specific legacy of being in social housing for both men and women at ages 23 and 33; ! Childhood behaviour indicators most specifically relate to adult malaise.
