5 Investing in the social sector
Social interventions cover a wide range of initiatives from regulation to the provision and funding of services to individuals. While there are many potential social initiatives that could contribute to well-being, resources are limited and costly, and choices must be made between them. Allocating expenditure on social policy and services is thus an economic problem, since it competes with other activities for scarce resources. The essence of the analytical framework is to consider social expenditure as an investment designed to improve aggregate and population sub-group well-being, with current costs and uncertain future benefits (Ministry of Social Development 2001). Viewing social expenditures as investments allows all the overall costs to be arrayed against the all the benefits over time so they can be compared in terms of cost-effectiveness, allowing resources to be shifted between them at the margin.
The design and delivery of effective interventions requires sound theory about the causal linkages between childhood and well-being in adulthood and an understanding of the impact of environmental influences on well-being. Good information is also needed about the nature, extent and cost of poor outcomes in childhood and adulthood. Evidence on the nature, size, distribution and timing of the effects of interventions and their overall costs is essential to determining the net benefits of social investments.
Understanding current and past initiatives can provide guidance when designing new ones. It is important not just to know if an intervention works, but also how and why it works and for whom. Knowing what actually happened in an intervention programme and under what conditions is vital to designing and adapting programmes to other contexts. Knowing why something works can also help in understanding the causal and mediating variables that contribute to an outcome (Boaz, Ashby and Young 2002). The causal nature of a targeted factor can be demonstrated through the intervention effect size (Huffman et al 2000b). The evaluation of interventions to determine their cost-effectiveness and the reasons underlying success or failure thus becomes an important part of managing the social investment portfolio.
The analysis can also be useful in designing strategies that achieve the desired goal at the least cost. It can highlight the extent to which alternative strategies can lead to savings. In this way, it can help decision makers to select among a range of effective interventions those that are the most cost-effective. For example, a study in Washington State involved the systematic analysis of a wide range of approaches to reduce crime, –from prevention programmes designed for young children to correctional interventions for juvenile and adult offenders – to help decision-makers allocate resources toward economically successful programmes and away from unsuccessful programmes, thereby producing savings(Aos, Phipps, Barnoski and Lieb 2001).
A portfolio approach could potentially be used to select an optimal suite of interventions in the social sector based on the quantification of the expected net present value (NPV) of the benefits, including estimates of the risk. In practice however, while quantification might be feasible for a subset of investments, it is unlikely to be applied across the whole sector. While some areas of investment might be amenable to quantification, the practical difficulties involved in quantifying in money terms the NPV of investments in other areas are likely to be formidable. A full cost-benefit analysis that also incorporated estimates of changes in the NPV or rate of return with changes in investment would be most useful in informing decisions about resource allocation. However, in practice data limitations may require other forms of evaluation to be used. Even so, the key to the investment approach is that both benefits and costs should be included in assessing the effectiveness of interventions.
There is considerable uncertainty about how best to improve well-being. The underlying causal relationships are not always fully understood, there is uncertainty about targeting and initiatives may not be fully effective. Such uncertainty suggests that a range of different approaches, focused at different stages and including universal and more targeted interventions, is likely to be appropriate to address particular outcomes. Addressing a particular outcome might thus involve a portfolio of different interventions.
5.1 Timing
Social policies and services are directed at different stages of a person’s life course, from pre-birth to adulthood. The opportunity to make investments at different stages raises a question about the optimal timing: is it better to provide services early or later in life?
Early interventions can be useful when the risk and protective factors are broadly understood, but there is limited understanding of the mechanisms by which these lead to particular outcomes (whether they are causal or not). By focusing on developing known protective factors (basic literacy skills, positive peer relations) or minimising known risk factors (child abuse), early interventions have a greater likelihood of being effective in reducing the probability of negative outcomes. Early interventions can prevent negative outcomes before they start to incur personal, social and economic costs and be more cost-effective because they can help prevent multiple negative outcomes.
However, the earlier the intervention, the looser the targeting is likely to be, the larger the potential target group and the greater the costs. Also, if the targeting is too precise, Type I errors (ie, missing those who should be targeted) are likely to be larger because problems may not be manifest at an early stage. It can be difficult to determine whether it was the provision of the intervention that helped prevent the negative outcome, whether the effect was a random one, or whether some individuals were exposed to factors other than the programme that may have improved their outcomes. There is also a risk that other intervening factors will undo the benefits of the early intervention (eg, poor quality schooling can undo the positive effects of early childhood programmes).
While it may seem sensible to intervene as early as possible to prevent the establishment of negative developmental pathways, there is some evidence to suggest that particular interventions work best when they are delivered at a critical point in time, such as a key developmental transition. Hence, decisions about the timing of interventions are not just a matter of early versus later. Rather it may be necessary to intervene at a point when they are most likely to be effective. Interventions before or after the sensitive period may be ineffective.
5.2 Success factors
Many factors influence the probability that an intervention will succeed or fail in improving well-being. One is the technical capability of the intervention to achieve the intended outcome. Technical feasibility is greater when the intervention is informed by evidence of causality, when the mediating factors being addressed affect the outcome of focus and when there is evidence of effectiveness. The quality of implementation is critical to the success of interventions (Derzon 1999).
Implementation issues include the quality of management, the clarity of aims and objectives, a receptive environment, effective leadership, the qualifications, training and sense of ownership amongst those delivering the programme and the level of resourcing. This indicates the importance of designing these factors into the intervention and of monitoring the progress and quality of implementation at the local level. A key issue in considering implementation is whether the government is better placed than other organisations to intervene, and if so, the best form the intervention might take.
Most of the empirical evidence is derived from the United States, Canada and the United Kingdom and a range of contextual factors needs to be taken into account in applying inferences from this evidence to New Zealand. While many elements of society, culture and the economy are broadly shared across developed countries, specifics can differ, for example in the extent to which disadvantaged families are concentrated in particular communities. Likewise, cultural values may differ across countries, and among different groups (for instance ethnic groups) within countries, and these values may have an important influence on mediating effects on children’s outcomes. Programmes with an apparently similar design may be successful in some locations, but not in others (Scott 2000). Imported programmes need to be underpinned by a good grasp of the context in which the programme was initially developed, and how this differs from the context in which it is intended to introduce it.
