3.6 Other findings from longitudinal studies

In seeking to understand the trajectory that leads to adult criminality, the longitudinal studies referred to above offer important additional insights. These studies, along with a large amount of overseas research, reveal a distinct pathway or “trajectory” that consistently increases the risk of offending by young people. Both studies followed large cohorts of children - over 1000 children - born within a certain area during a set time period. It is acknowledged that these studies can be criticised because Māori were to an extent under-represented in each (12% of the Christchurch study, and 11.3% of the Dunedin study) and the local social context is not necessarily typical of New Zealand as a whole. However, their value with respect to illuminating childhood and adolescent development, with respect to offending behaviour, is unique.

These two studies have had different but complementary emphases. The Dunedin study tends to examine the psychology of the person, and the Christchurch study tends to focus on the environment. This complementarity is useful for considering the range of outcomes from the interactions of individuals with their environments.

A unique strength of longitudinal studies of this type is their capacity to identify causal relationships, rather than simple correlations or associations. This requires sophisticated statistical analysis, whereby the influence of various factors, which might potentially bear a causal relationship to an outcome, can be selectively “controlled for”, allowing the genuinely causal influences to emerge (similarly as is described above in relation to the Department’s Home Detention study).

A good illustration of this process emerges from the data from the Christchurch study, in relation to Māori involvement with violent offending. Analysis shows that what can at first appear to be a clear and significant relationship between violent offending and being Māori, is in fact more complex, and inseparable from the family background factors described earlier. Table 6 below shows that Māori in the Christchurch longitudinal study were more likely than non-Māori to have committed a single violent offence, or repeated violent offending, and/or to have a Police record for violence. Statistical analysis showed that the probability of these differences happening by chance is extremely low, and Māori appeared to be involved in violent offending at two to three times the rate of non-Māori.

Table 6: Violent offending and involvement in violence among Māori and non-Māori study participants at age 18 (%)

Māori (n=96)

Non-Māori (n=929)

Relative risk


Any violent offence  (17-18 years)





Repeated (3+) violent offences (17-18 years)





Police record for violence (ever)





Source: Christchurch Health and Development Study

However, when the specific influences of social and family factors were examined, ethnicity of the young person ceased to function as a significant explanatory variable. In other words, family circumstances (parents’ history of problems with alcohol, family history of offending, multiple family problems) and parenting behaviours (use of physical punishment, level of parental care) emerged as the key factors which accounted for the violence in these adolescents 1. Wherever such factors were recorded in a family - irrespective of ethnicity, or indeed even socioeconomic status - the risk of offending amongst children in those families increased sharply. Such adverse factors often occur together and have a cumulative effect. Section 6 of the Appendix reproduces graphs from analysis of the Christchurch study which illustrate some of the interactions between socioeconomic factors and deviant and criminal behaviour.

The Christchurch study concluded that ethnic differences in violent behaviour were not significant when family and developmental factors were accounted for. The problem appears to be that, for a range of reasons, many Māori young people appear more likely to be exposed to such adversity in their home environments. While Māori are certainly more likely to face disadvantage in terms of socioeconomic status, housing, and education, disadvantages of this type, in and of themselves, do not seem to account for differences in outcomes. What these studies indicate therefore is that many contemporary Māori families are beset by higher levels of family dysfunction and difficulties than are non-Māori 2. In other words, those life circumstances most often associated with offending are, for a range of reasons, more likely to affect Māori families.

Fergusson’s own summary is as follows:

(I)n broad outline it seems likely that the difficulties and disadvantages faced by contemporary Māori families are likely to represent the end of a long term historical process that has involved many components, including: the pressures faced by, and change in Māori culture and language following colonisation, the loss of land and economic power base experienced by Māori, increasing urbanisation of Māori and the general reduction of status and prestige (mana) of Māori people within the context of New Zealand society.

1 Fergusson DM. Ethnicity and interpersonal violence in a New Zealand birth cohort. (2003c) in Hawkins, Darnell F. (Ed.) Violent Crimes: Assessing Race and Ethnic Differences. Cambridge; Cambridge University Press, 138-153.

2 Fergusson (2003c) op. cit.