Social services include a wide-ranging mix of income transfers, care and protection services, developmental activities and remedial actions.
Despite the resources used, and the potential for both harm and good that can be done, we often do not have an analytical basis available for determining why most programmes take the form that they do, and how well they work.
Some of this is because of the nature of social services while much is because the past choice of social services mainly reflects elements of social and cultural values and preferences that make them immensely political. Many social services involve partly connected networks of groups and agencies, often operating with some form of contract to a government agency, with partial information about the people that they are helping, and with separate and potentially different ways of defining people and their conditions, and establishing needs.
There are fundamental elements of social services that result from politically determined rather than scientifically based perspectives on the nature of programmes: universal versus targeted, service provision versus cash, free or part charges, outsourced or public provision, and the nature of the emphasis on human rights. Philosophical attitudes to taxation, regulation, penalties and sanctions also shape political preferences. Certainty in the means can occur without clarity in the results sought.
Whatever the political perspective, evidence is essential to improve the quality and robustness of service outcomes over the time periods for which costs and benefits are to be compared, as well as the ability to take account of the relative impact of dispersed contributors to improved outcomes and of the breadth of outcomes considered. Even so, the selection of social services is primarily a political action, and the imperatives of deciding what to do will rarely await the discovery or implementation of new information sources.
At its crudest, programmes will be judged by how much funding will be made available for them rather than what they will achieve, as typically occurs at the time of general elections. In the most recent general election in New Zealand (2017), policy commitments were generally expressed as ‘how much’ rather than ‘how useful’.
One consequence of serious concern is that once decisions have been taken, the political risk that previously unavailable information might adversely challenge the original decision can result in the prevention of improved information sources that might better inform future situations, and the quality of services that are provided. The expectations on social policy and service delivery are continually being extended.
The rise in importance of a social concern generally occurs without any formal process, and often it is the community sector that triggers interest in concerns that get addressed at a government level. Issues that are currently in various emergent stages of wider recognition in the policy information system include: obesity, social media and technology, super diversity, suicide, harm and violence, abuse, bullying, mental health, drugs, pornography, antibiotic resistance, homelessness, crowding, fertility, urban infrastructure, incarceration costs and third world diseases.
Where we can develop an evidence base to inform a political or institutional response, the long-term viability and impact of that response on citizens will reflect the limits to measurement and analysis. Without knowledge of the evidence base there is likely to be unjustified confidence in services, and this can bring damaging consequences for groups of citizens.
In the social services, institutions and politicians appear more likely than in other sectors of public administration to have a strong aversion to evaluation and continuous improvement practices that make transparent the imperfection inherent in their decisions and complicate managing political risk. There are many examples where the worlds of science, policy and practice remain not as readily connected in the social services as might be presumed. It is not unusual for long-standing social services to have been at best poorly tested and evaluated, with the consequence that the final form of many programmes is not based on relevant evidence or regularly tested by replication, pilot studies or continuous improvement practices.
Observational evidence always competes with anecdote, belief, un-validated theory or just prejudice, not just in setting policy but also in determining whether to continue to gather evidence or invest in new forms. Where evidence is underused through choice, limited competence or underinvestment, then anecdote, belief, un-validated theory or prejudice are more likely to drive decisions. The very existence of evidence that people have reason to trust can inform or challenge the political or institutional preferences that frame thinking and problem solving.
There is an inevitable and potentially quantifiable uncertainty and system risk inherent in social services which are poorly understood by the public and by politicians, for example in decisions such as granting parole. Where we have little idea of the quality of the evidence used to justify the policy, we cannot take account of that uncertainty in managing the delivery of the service, such as the undoubted lack of precision in any screening criteria. Consequently, with many social services we are often unable to be certain about what happens for the people who have a need and entitlement for them, and why only some of these people connect effectively with a service that they are eligible to obtain.
 “Three strikes and you’re out” example, and boot camp proposals