10 March 2007

Targets and experiments

Commenting on the educational ranking devised by the UK’s Department for Education and Skills (DfES), the [London] Financial Times points out that:

Secondary schools that encourage their pupils to sit non-traditional subjects such as leisure studies and performing arts are bolstering their league table position at the expense of leading schools that focus on core disciplines, a unique analysis of exam results has shown. … Lancaster High School, a comprehensive ranked 478th by the DfES plummets more than a thousand places to 1,623 when “softer” subjects and vocational qualifications are excluded.
This is but one illustration of how the use of badly thought out numerical indicators and targets influence behaviour for the worse. They point to a genuine problem for well-meaning policymakers whose remit covers any but the smallest community: which numbers to use and how to use them? The Social Policy Bond approach relies totally on numerical targets and so, increasingly, does current policymaking, for better or worse.

It seems to me that useful indicators and targets should be inextricably correlated with wellbeing. And we should be targeting the wellbeing of ordinary people – not corporations or institutions, which have entirely different goals that can and often do conflict with those of the public. Importantly too, our chosen targets should also be expressed in broad terms, mainly so that objectives that can conflict with each other are targeted by a single policy – or issue of Social Policy Bonds. This is in order to minimise the chances that people will achieve one targeted goal at the expense of another, untargeted component of wellbeing. Broad objectives also maximise efficiency gains, as they give wider scope for the deployment and shifting of scarce resources.

The need for broad targets highlights one of the difficulties I experience when trying to raise interest in the Social Policy Bond concept: they function best over a large scale, making small-scale experiments difficult to devise. Ideal policy areas for Social Policy Bonds are those where a large range of approaches and projects need to be explored, adapted, tried and (if unsuccessful) terminated or (if successful) adopted widely. Climate change is one example. Others are eradicating poverty, raising literacy, reducing crime rates, over large regions or whole countries. It’s difficult to devise experiments at that level, but an alternative might be a virtual experiment in a simulated world, such as at http://www.secondlife.com/.

No comments: