7 Conclusion
There is increasing concern by governments that they should maximise the benefits arising from their expenditures. At present however, there is no systematic way of allocating scarce resources in a way that enhances overall well-being. Current allocations reflect implicit judgements about the expected value of the expenditures. However, there is no clear means of identifying the outcomes of interest, how interventions can contribute to the outcomes or how cost-effective they are. Identifying the outcomes that matter and the interventions that are most cost-effective in achieving them can provide information to decision-makers about the best ways to deliver results.
The central aim of this paper was to provide an empirically-robust framework that can be used to compare intervention across a range of social sectors. There are two key components to the framework.
The first is a life-course view of child development that emphasises that experiences and influences in childhood can affect well-being throughout life. This approach involves investigating which factors contribute most to well-being, and which factors lead to negative outcomes. Understanding the underlying causes of good and bad outcomes can be helpful in designing effective interventions that enhance well-being.
The second component involves viewing social expenditures as investments addressed at achieving particular outcomes, typically directed at enhancing well-being. This approach provides a means of comparing the costs and benefits of interventions of different types directed at specific outcomes. A social investment perspective, whether quantitative or qualitative, involves linking interventions to their intended outcomes by seeing how well they modify the causes of poor adult well-being; considering the size and distribution (across people and time) of the costs and benefits; and assessing the extent to which the potential benefits can realistically be achieved.
The evidence indicates that child development is the consequence of a large number of different influences, each of which, with the exception of individual heritable factors, plays a relatively small part in the eventual level and pattern of adult outcomes. This suggests that programmes and policies that attempt to target a single adverse influence or risk factor are unlikely to be effective. As Fergusson (1998, p 171) comments:
What the CHDS findings clearly suggest is the need for policies that are targeted at meeting the needs of at-risk families and at-risk children. This theme has clearly emerged in the analyses of multiple-problem adolescents and suicidal youth that showed both of these endpoints were determined by combinations of risk factors that conspired over the individual’s life course to increase the risks of personal maladjustment or suicidal behaviour.
Furthermore, the evidence shows that a considerable part of the variation in child and adult outcomes cannot currently be explained in a secure causal sense. There is considerable remaining uncertainty about the nature and strength of the causal linkages that act through time and no overarching theory to tie together the research findings. This suggests that it would be extremely unwise to place undue reliance on any one form of intervention or response to poor childhood and early adult outcomes or to put too much emphasis on any one stage in the developmental path between conception and mature adulthood.
This paper has reviewed the available evidence on the effectiveness of interventions to address selected outcomes: youth suicide, teenage pregnancy, educational underachievement and youth inactivity. Most of the evaluative evidence comes from the USA and relates to discrete programmes and interventions, usually targeted at children or young people defined as at-risk in some way, or focused on the needs of individuals who have already demonstrated problematic behaviour or experienced poor outcomes previously. Even then, the evidence is very patchy. Some areas have a large number of high-quality studies and have good meta-analyses and systematic reviews; others have a far weaker empirical base. In other areas, there is evidence, but it fails to provide clear guidance for policy. For example, the evidence on effective programmes to reduce teenage pregnancy – of concern in New Zealand (though with less causal significance for later outcomes than often thought) – fails to give a clear indication of what works. The difficulty of establishing a clear sense of what might be effective may reflect the complexity of the underlying causes of teenage pregnancy. If causes operate at different levels, then the pattern of intervention should too. Studies of individual interventions cannot accommodate this easily.
It is also apparent that the number of good studies concerned with relative cost-effectiveness of different programmes and interventions is far smaller than those focused exclusively on effectiveness. Universal programmes, which consume the majority of public resources, are far less likely to be evaluated than targeted programmes, partly because they are much more difficult to study with a comparative or experimental design and partly because they are likely to be far more difficult to alter significantly than more marginal, selective programmes. Some types of intervention are rarely considered in evaluative research terms at all. Most legislation affecting children and families fits this category. Much more needs to be done to identify the potential consequences of legislation. This has not been possible in this paper.
A number of different types of general interventions that aim to enable children to deal with a wide range of adverse circumstances and are focused on at-risk families with young children have been shown to be effective at reducing the risk of serious problems arising in later life. In later childhood, interventions tend to be less concerned with health and general psycho-social development and more focused on reducing the risk of specific adverse outcomes. Interventions designed to reduce risk-taking behaviours among adolescents and young adults tend to focus on substance abuse, sexual activity and driving.
Given the importance of education as a protective factor against later adverse outcomes, such as criminal behaviour, raising the educational achievement of those who are likely otherwise to acquire the fewest skills is a priority. The most effective way of doing this appears to be through early interventions with at-risk children focused on school readiness, literacy and numeracy. There is also evidence for schooling interventions, both systemic and pedagogical, that are effective in raising cognitive skills and educational attainment.
The evidence on effectiveness and cost-effectiveness provides some guidance for implementing a cost-effective portfolio of policies designed to improve children’s and young people’s well-being. Interventions should be assessed, not in isolation, but as elements in a balanced portfolio of investments to improve well-being through the life-course. The portfolio should contain a number of different approaches to addressing the different outcomes given the uncertainty about causation; a mix of early and late interventions; targeted and universal programmes; and interventions focused on different outcome domains relevant to eventual adult well-being.
Given the existence of major universal programmes in areas such as health and education, improvements in outcomes are likely to be brought about at the margin by either making adjustments to these programmes with a small effect on a large number of people and/or by targeting individuals and groups particularly likely to experience poor outcomes in later life using relatively intensive interventions.
Particularly in the early years, a mix of universal and selective interventions is likely to be needed because poor outcomes are not always visible, early risk factors do not explain all later outcomes (Type I and Type II errors) and universal programmes can provide a basis for screening the population for more targeted interventions. Later on, more selective programmes are likely to be more effective for those at risk of poor outcomes, though this can come at greatly increased cost.
The design and selection of interventions should reflect the fact that many causal factors may have their effect by acting in combination, so that a portfolio of interventions that between them address multiple risk factors (ie, by acting at different levels such as the individual, family, school and community) is likely to be most effective.
The main drawback of preventive interventions is the likely long time-lag between the intervention and its effects in terms of outcomes such as reduced offending, though well-targeted early childhood home visiting programmes can be cost-saving to government when later delinquency and crime costs are computed. This suggests the need for a mix of early preventive programmes and later programmes with children who have already shown some signs of antisocial behaviour, both targeted on those most at risk.
It is important to base the resulting portfolio of investment in children and young people on secure evidence of what works, for whom, at what period of development, for how long and at what cost. However, implementation is a vital part of effective interventions and needs to reflect particular contexts and cultures. Simply copying interventions that may work overseas is not sufficient to ensure that they are effective in New Zealand. Also, it is important to consider not only when government intervention might be superior to alternatives, but also whether regulation, service provision or purchasing is the most appropriate form of intervention.
Effective interventions should be designed and selected cognisant of differences between age groups in patterns of development, the dimensions and causes of well-being and the need for different emphases and strategies at different ages. Some actions designed to improve well-being will address a specific age group while others may benefit the entire period from infancy to adulthood. Different groups need different degrees of intervention ranging from normal access to the “core” universal programmes (eg, compulsory schooling) to enrolment in targeted, comprehensive, long-term programmes.
This paper represents a preliminary overview of issues at the heart of the well-being of children and young adults in New Zealand. It has only begun to bring together the evidence and insights necessary to operationalise a social investment framework. Much remains to be done to strengthen and deepen understanding of the causal relationships between childhood and adulthood, and, in particular, the factors underlying poor states of well-being in New Zealand; and the evidence from both New Zealand and elsewhere of the effectiveness of interventions. The paper does not include the detailed description and understanding of the current New Zealand configuration of investment, which forms the point of departure for any analysis of the relative priority to be given to different sorts of investment in light of the evidence of the likely costs and benefits of different investments. This analysis forms a central part of future research.
