How many policy announcements will be evidence-based during party conference season

1 September 2020

The month of September is filled with policy promises from across the political spectrum. Some of these will sound like vote winners, but in reality, not many will have been tested in any meaningful way.

Science tests its theories. Companies have to prove safety or efficacy before bringing certain products to market, think of medical or chemical trials, why shouldn’t public policy be the same?

Theory and practice can be worlds apart – even if the theory sounds irreproachably sensible. Take a simple example of behavioural economics such as fines introduced for late pick up from a childcare nursery.

In theory, a financial penalty should deter bad behaviour, but in fact the opposite was revealed. Parents treated the fine as a price that justified their late arrival. The nurseries that introduced the fine found twice as many parents were late. Guilt turned out to be a more powerful incentive than money. A charge simply allowed users to push the boundaries.

Experimental methods, driven by plentiful data, are a common feature of commercially driven insights, but now need to become standard in the public policy-making context. Nesta, the UK innovation think tank is leading the way through its Innovation Growth Lab which is providing some excellent case studies, but progress across government is slow and far too piecemeal.

The deepest policy problems and questions – what would improve our economic productivity for example – need far more robust research methods including longer pilots and randomised control trials, not always a welcome reality if you are a politician in search of an announcement.


Tags: Big Data, Data Analytics