Better public services through experimental government

This paper argues for the adoption of an experimental ethos by government would result in better social policies and greater value for money.

Initially targeting areas where more innovation is merited and politically feasible, the paper sets out recommendations for how to foster experimental government.

Key findings

  • A lack of evidence and testing to find out what works means that social policies are being blindly rolled out by governments.
  • Adopting an experimental ethos would mean that governments could take a systematic approach to trying things out and see what works before rolling policies out more widely.
  • An experimental approach to government still leaves room for values and beliefs – which should always be at the core of government – while creating room for experimentation if not in policy formation then in its delivery.

How do we know, in advance, that a social policy will work? Often the honest answer is we don’t because, unlike other areas where experimentation is gaining currency, in social policy, we don’t test things out. This paper advocates for the need for an experimental ethos in government. Arguing that testing things out doesn't mean the abandonment of values and beliefs. Rather it allows for the systematic testing out of policies before roll out thereby supporting the efficient use of public funds and the maximising of public good.

Policy recommendations

Several changes need to take place to facilitate the adoption of an experimental approach to government. Initially targeting areas where more innovation is merited and politically feasible, we recommend that:

  • more ambitious and bold experiments are set up on nationally important issues
  • experiments use the most robust evaluation methods available
  • leaders in government embrace risk and reward success
  • the right institutional support for experimentation be supported
  • researchers and government officials work closely together to co-produce experiments
  • opportunities for experimentation be seized
  • a ‘learn as you go’ philosophy is adopted, one that does not presume policies to be set in stone
  • public debate on the importance of experimentation is encourage


Jonathan Breckon


Jonathan Breckon

Jonathan Breckon

Jonathan Breckon

Director, Alliance for Useful Evidence

Jonathan was the Director of the Alliance for Useful Evidence from 2012 to 2021.

View profile