2.3  Review methods

Section 3 of this review looks at the relationship between educational attainment and outcomes over a range of different areas of life, considering each in turn, and investigating whether there is evidence that increased education causes good outcomes.

Publications were found by searching a variety of databases and on-line collections, including the EconLit database of economic literature, the Medline database of health literature, the ERIC database of education literature, the National Bureau of Economic Research (NBER) website, the full list of Christchurch HDS and Dunedin MHDS publications, and websites of longitudinal surveys in the United States and United Kingdom. Searches were undertaken using the terms “education” and “schooling” in combination with topic areas such as “health”, “mortality” and “social capital”, and with keywords such as “wider benefits”, “social benefits” and “non-market benefits”. Most literature was found, however, through following up references in publications.

In the areas of substance use, mental health, crime, welfare dependence, fertility and sexual risk-taking, the main sources of information on possible relationships with education were the two New Zealand longitudinal birth-cohort studies: the Christchurch HDS and the Dunedin MHDS. These studies are particularly rich in information on the participants’ early lives, school careers, and outcomes in early adulthood (participants are currently in their mid-to-late 20s). Because of this wealth of information on potentially confounding factors, and because they are studies of young people in New Zealand, the New Zealand birth-cohort studies are accorded particular weight in this review. The results from these studies are tested however, where possible, using evidence from studies using different methodologies or which have been undertaken in different countries.

Outcomes which typically occur later in adulthood have not yet been investigated in the New Zealand birth-cohort studies but have received attention in various overseas studies. Section 3 includes discussions of the relationship between a person’s education and, variously, their physical health, social connectedness, political participation, and the education of their children. Finally, the relationship between a person’s education and an overall measure of benefit, namely their subjective quality of life, is discussed.

Where possible, only the results of longitudinal studies, IV studies or twin and sibling studies are used in this review, for the methodological reasons discussed above. If there was no better alternative, however, then other, less rigorous studies were used, and this is identified in the discussions. Again where possible, only the results of studies which were specifically set up to look at a relationship with education are used.

It is important when interpreting the results of the studies reviewed in this paper to note that an increased, or better, education is measured in these studies in only two ways. One is by measuring the duration of education: that is, by recording the number of years that a person spent in formal education, or the age at which they completed their education. The second is by measuring their educational attainment: that is, by recording the highest qualification they held, or their attainment of particular qualifications (whether or not, for example, they passed School Certificate). Therefore studies do not look, for example, at the effects of going to a better-funded school, having better teachers or being in a smaller class. The effects of changes in school quality have been investigated to some extent in the literature on earnings-related benefits of education (e.g. Card and Krueger 1992), but not in the literature on wider benefits.

In addition, this review does not look at the potential consumption benefits of education (that is, enjoying or benefiting from education at the time that it is undertaken) and does not look into all the aspects of life which authors of other review papers have discussed. This review does not discuss, for example, non-wage job benefits; efficiency of labour market searching; productivity in the home; efficiency of consumer decision-making; charitable giving; higher rates of savings[7]; improved sorting in the marriage market; effects of education on one’s spouse; effects on one’s children (other than on the children’s education); attainment of desired family size; efficiency in contraceptive use; effects on the ageing process; effects on enjoyment of entertainment; democratisation, human rights and political stability; environmental equality; racial tolerance; support for authority; political cynicism; and membership of the local PTA.[8] This was in part because these areas are less well-studied than others. Also, to keep this review from becoming unmanageable, it was necessary to limit the number of aspects of life which were investigated. Any outcomes which are not specifically discussed should in any case be captured under the general category of ‘life satisfaction’ in section 3.9.