Skip to navigation Skip to main content

Five reasons why behavioural insight isn’t being used to improve public services

Jon Ainger


Like many hundreds of other eager and interested people, I attended the star-studded BX2015 conference a couple of weeks ago.  Packed with luminaries of the rapidly emerging field of behavioural science, it was filled with some of the most senior, experienced and recognised leaders in their fields – Nobel Laureates included.  Each session was packed with great insight and handy tips to take away.

And yet – despite the pioneering work of the Behavioural Insights Team, books such as Nudge, and the work of companies such as ourselves at IMPOWER, public services remain stubbornly resistant to the application of behavioural insights at any real scale.  Recently, we did a review of 10 A&E departments and found that, on average, 50% of the opportunities to use behavioural science to reduce demand were being missed. Worse still; of the other 50% a big proportion were having the opposite effect – driving up demand. Similarly, most local authority customer interfaces remain clunky, and uninformed by basic behavioural science.

Given the financial pressures on the public sector, what’s stopping us?  I’ve listed five reasons below:

  1. It’s surprisingly easy to learn, but surprisingly hard to apply.  The various models (all of which have the same basic components) are packaged up nicely, with easy to remember acronyms (remember – these are behavioural scientists!)  But putting this stuff into practice and showing it works takes some bravery and some persistence.  (And someone with a budget has to spend money on it.)
  1. Different data is required.  Whilst it’s easy to say ‘I want less littering’, it’s much hard to answer the questions ‘by which people, when, and what type of littering?’ Answering these questions requires different data sets (data which is often non-trivial to secure).
  1. Trials require change, and are easy to damage.  We at IMPOWER are not slaves to the Randomised Control Trial approach, but the basic point that we should measure whether something works before rolling it out is unarguable.  The problem is that a) trials often rely on people working differently and b) everyone else around them is still working in the old way as a control.  You can use smart trial design; but the lack of compliance of one or two members of staff can ruin a trial.
  1. System opportunities are sadly underexploited.  The public sector is an inter-connected, large, complex, system.  Behaviour in one part causes costs in another.  We have shown it is possible to change the costly behaviour of partners (e.g. re-referring children into social care) through behavioural interventions. But often, this isn’t seen as core business and goes missing as an opportunity for improvement.
  1. It’s still a bit weird.  Okay – it’s cool at the moment.  But at the heart of behavioural science is a basic message that we should not bother to change people’s minds.  Instead, we should change their behaviour.  This just feels wrong to many intelligent, logical, analytical people – many of whom work in public services – who have made their careers on being smart and persuasive.  And it’s hard to quit.

Overcoming these barriers isn’t impossible.  More and more trials are being carried out in the public sector, by us and by others.  But in an ironic twist, until it becomes ‘normal’, we will continue to miss opportunities to save money using behavioural science.

To find out more about how to apply behavioural science to improve public services, get in touch.

Twitter: @jonainger

Written by

Jon Ainger

Director, IMPOWER



Sign up for the latest thinking on delivering sustainable change and better public services

No spam; unsubscribe easily at any time. Learn more in our Privacy Policy.