3  Evidence on each outcome

3.1  Substance use

Of the Christchurch HDS cohort, 19% left high school without any formal qualifications: that is, they had not achieved at least one C grade in a School Certificate subject. Nationally, about the same proportion of students leave school with no qualifications (Ministry of Education 2002b).[9]

Using the Christchurch HDS, Fergusson, Swain-Campbell and Horwood (2002) investigate the association between leaving school without qualifications and clinically-defined dependence on alcohol, cigarettes and illicit drugs between the ages of 18 and 21.[10] To account for factors occurring before the participants sat School Certificate, which might confound the relationship between education and substance use, Fergusson et al control for a wide range of background variables. In particular, they control for gender, for deviant behaviours and conduct disorders in childhood, for parents’ use of drugs or alcohol, and for level of family disadvantage. Each of these factors is believed to put young people at greater risk of alcohol and drug abuse (Stewart 1997; Chetwynd 1997). Fergusson et al also control for prior cigarette smoking – although not prior alcohol or drug use – and for natural ability by means of an IQ test in childhood and a test of general scholastic ability at age 13.

Fergusson et al find no persistent link between leaving school without qualifications and dependence on alcohol or drugs. There was, however, a persistent link between school qualifications and nicotine dependence. Other studies have also shown relationships between education and smoking. Hu, Lin and Keeler (1998), for example, in a study of Californian teenagers, find that young people who did better at school were less likely to take up smoking, and more likely to try to quit, even after controlling for other socio-demographic factors.

Mensch and Kandel (1988) look at the connection between alcohol and drug use and dropping out of high school in the United States. To complete high school in the United States requires staying at school until approximately age 18. Mensch and Kandel use data from the National Longitudinal Survey of Youth (NLSY), a survey of over 12,000 young people, who were aged 14-21 when first interviewed. They find a higher use of cigarettes and illicit drugs, but not alcohol, amongst high school drop-outs, after adjusting for socioeconomic, behavioural and personality factors known to be related to early school leaving. Their interest, however, is in determining whether drug use contributes to dropping out of school, rather than whether dropping out contributes to drug use. By looking at the age at which drug use was initiated, Mensch and Kandel find that the use of cigarettes, marijuana and other illicit drugs predicts early school leaving. Furthermore, the younger the initiation into drugs the greater the likelihood of a student leaving school without a high school diploma.

It appears, therefore, that young people who are at risk of later alcohol and drug use also tend to leave school without qualifications, and this explains the apparent association between educational achievement and later drug use. Some drug use during adolescence may in fact make it more likely that users leave school without qualifications. The exception may be for cigarette smoking, where Fergusson et al, along with other studies, find a persistent link between education and later smoking habits. Fergusson et al consider that smoking may have become an accepted pastime for young people who are not working or participating in further study or training. The link between education and reduced smoking rates may also be due to greater understanding of the health risks of smoking (Sander 1995a; 1995b).

3.2  Mental health

From the Dunedin MHDS, Miech, Caspi, Moffit, Entner Wright and Silva (1999) study the relationship between educational attainment and mental health disorders at age 21. Since mental disorders are potentially both a cause and a consequence of poor educational performance, Miech et al also take account of pre-existing disorders at age 15. The mental disorders they consider at both ages are anxiety disorders, depression, and antisocial disorders (that is, a pattern of disregard for, and violation of, the rights of others).[11] Educational attainment was measured by passing School Certificate and Sixth Form Certificate, and by attendance at university.

At age 21, study members with low educational attainment reported significantly higher levels of anxiety, a finding which persisted after controlling for anxiety at age 15, parental socio-economic status and gender. The relationship between educational attainment and anxiety was monotonic: increases in anxiety between ages 15 and 21 were highest amongst those with no school qualifications and progressively declined with higher educational attainment.

Study members with low educational attainment were also more likely to meet clinical criteria for conduct disorder and antisocial personality disorder, after controlling for conduct disorder at age 15, parental socio-economic status and gender. As with anxiety, the relationship between antisocial behaviour and education was monotonic. Increases in antisocial behaviour were highest amongst those with no school qualifications and lowest amongst those who had been to university. There was no significant relationship, however, between depression at age 21 and educational attainment. Furthermore, Miech et al found that depression at age 15 did not appear to influence subsequent educational attainment.

From the Christchurch HDS, Fergusson, Beautrais and Horwood (2003a) identify factors which promote vulnerability and resilience to suicide amongst young people with major depression. Overall, 38% of the sample met clinical criteria for major depression at least once during the interval from 14-21 years; 29% reported having suicidal thoughts at some time during this period; and 7.5% reported having made a suicide attempt. Most of the suicide attempts were not medically serious, however. From an initially wide range of individual, family, peer and school factors, Fergusson et al construct a reduced-form model of significant risk factors affecting the likelihood of suicide attempts. These results show that rates of suicide attempt, but not suicidal thoughts, were higher amongst those young people with fewer School Certificate passes. Rates of suicide attempt were also higher amongst young people with depression, who had a family history of suicidal behaviour, who had been sexually abused as children, who had deviant peer affiliations, and whose personality tended towards novelty seeking and neuroticism.

In another South Island study, although in this instance a case-control study rather than a longitudinal study, Beautrais (2003) compares three groups of young people: a group of 60 who had committed suicide, a group of 125 who had made serious suicide attempts requiring hospital treatment, and a randomly-selected control group of 151. Comparisons were made on a series of measures, including socio-demographic, childhood, family, psychiatric, and psychosocial factors. Information was collected from each participant’s ‘significant other’. Poor school performance was found to be a significant risk factor for suicidal behaviour. Young people who had committed suicide were much more likely than people in the control group to have left school with no qualifications, even after controlling for all the background factors. Young people who had made a serious suicide attempt were also much more likely to have left school with no qualifications.

The findings from these studies suggest that poor school performance, and, in particular, leaving school without qualifications, may increase anxiety disorders and foster antisocial disorders in young people. It may also increase a young person’s risk of attempting, and also completing, suicide. Poor school performance does not appear, however, to increase the likelihood of depression.