Policy Lunchbox: Is experimental Government possible?


As research scientists, we’re used to experimenting. Spending time at the lab bench kitted out in a lab coat and safety glasses is our natural habitat. But scientists aren’t the only ones who experiment. It seems that the Government is at it too.

Our latest Policy Lunchbox event featured Jonathan Breckon, Head of the Alliance for Useful Evidence, who spoke on the topic ‘The What Works Centres: Is experimental Government possible?’

The What Works initiative involves a network of centres set up by the Government. These aim to use evidence to make better decisions to improve public services. Six of these centres were initially set up and now operate at arms’ length from the Government. The centres cover a wide range of areas from effective policing to aging to local economic growth. Arguably the most famous former What Works centre is NICE (the National Institute for Health and Care Excellence). Both Scotland and Wales have specific centres; unusually the latter is not independent from the Welsh Government and gives direct advice. There are a number of new centres planned including one which focuses on wellbeing to be led by Public Health England.

These centres serve to generate evidence in a usable function, but is there the capacity to use it? How do we ensure that there is a demand to match the supply? Jonathan argued that experimental Government, while challenging, is possible.

He argued that we need our Government to be an experimental and learning one that proceeds by testing and evaluating new ideas. The Government wouldn’t do the research itself but would use the best and most robust evidence available to test out benefits and problems first before implementation.

Examples of this is practice exist internationally. The progressor project in Mexico is one such example and aims to reduce poverty. Conditional cash payments are made to the poor if they sign up to education and health help. This began with randomised control trials in different areas and has been rolled out to approximately 5 million people since 1997.

In the UK, a number of initiatives have brought us close to experimental Government. A good example is the Behavioural Insights Team, the so-called ‘nudge unit’, which has trialled out new ways to increase sign up for organ donation. A range of approaches were tested, including adding a sign-up box at the end of completed forms (e.g. after the dreaded tax return) to capitalise on people’s good feeling after having completed an often dreaded task. In this way, approximately 350, 000 extra donations have been pledged at low cost.

However an experimental Government requires a degree of humility which is not always compatible with politics, particularly in the UK where a U-turn can be viewed as political suicide. In order for experimentation to be truly effective, there must be the capacity to address when something isn’t working. We need to allow politicians to make mistakes and learn from them. We need them to be able to adapt and move on for the sake of society.

This culture is more prevalent in the US. For example, an initiative to reduce crime by getting children to meet convicts was trialled, but evaluations revealed a negative effect and the project was halted.

Of course, as with many scientific experiments, there are ethical considerations. Not least the issue of informed consent… for example how would you convince parents of the benefits of their child participating in a school-based trial? We need experimentee willingness and an electorate which is happy with trials.

No easy feat.

So is experimental Government possible? Watch this space.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s