PM's nudge unit: why government needs to use randomised trials

article-body

For public servants who want to use behavioural economics, randomised controlled trials are a must. The PM’s nudge unit is working to build the capability.

The federal government’s behavioural economics team has produced a short guide to randomised controlled trials, which are essential for public servants who hope to subtly influence the behaviour of citizens.

They are one of two “core pillars” of the projects that are run by the Behavioural Economics Team of the Australian Government (BETA), the other being the design of the “behaviourally-informed interventions” themselves, according to a new guidebook written by Harvard University professor Michael Hiscox, who leads BETA, and adviser Phil Ames.

RCTs have been attached to the idea of influencing citizen behaviour by gently tweaking government communication materials all along, or at least since the White House first hired one of the academics who coined the term ‘nudge’ to describe it.

The United Kingdom’s behavioural economics team, which replicated itself inside the New South Wales Premier’s department, also strongly promotes small-scale experimental trials to test each new potential nudge against a control group. The NSW nudge unit recently produced a guide to some of the most common and well understood cognitive biases, which are another core concept in this increasingly popular modern approach to service delivery.

The PM’s nudge team says the use of behavioural economics is just another form of user-centred service design that is based on observing how large groups of people really behave rather than simply relying on incentives and disincentives to work on the basis we are all generally rational actors:

“Traditional policy makers assume people will always make the best decision possible, and have no shortage of willpower. However, research and evidence tells us this isn’t always the case.

There is often a gap between what people intend to do and what they actually end up doing. For example, when people are in ‘auto-pilot’ we know they will often use shortcuts and rely on biases and stereotypes to make decisions and, in some cases, people won’t act on their best intentions due to choice overload and complexity.”

It suggests cigarette plain packaging and pre-filled tax forms are examples of inexpensive nudges that have increased public service efficiency and encouraged people to “put their good intentions into action” by understanding how people typically behave. Other typical behavioural interventions involve changing the wording of correspondence or even the colour of the paper it is printed on.

In many cases, traditional incentives and disincentives — legal penalties, discounts, tax breaks and the like — can have very different effects depending on how they are communicated to the people they intend to influence.

Rachel Dixon, one of the project leaders brought in from the private sector by the Digital Transformation Office, also says digital service designers should observe how people really behave, rather than relying on their explicit views from surveys, for example.

BETA’s mission, the new guide explains, is “to build behavioural economics capability across the public service and drive its use in policy development and service delivery design” by running the trials in partnership with other federal agencies.

The small unit inside the Office of Best Practice Regulation also works to “establish links between the APS and the behavioural economics research and practitioner community, here and overseas” — and Hiscox should be able to open some doors in that regard.

Its projects start with a desired outcome and are structured in four stages: discovery, diagnosis, design and delivery. A good understanding of how to set up RCTs is mostly relevant to the first two, which involve identifying the problem to solve, researching the target population and proposing ways to nudge them towards the desired outcome.

The new guide from BETA walks the reader through the discovery and diagnosis phases of its model for APS nudge projects in a logical, step-by-step fashion, while an appendix provides an introduction to RCTs for those without a scientific background.