The most effective public policy and practice nudges of 2018

(QESP Editor’s Note: The following is a reprint of a 24/01/2019 article in The Mandarin. The original, with links to related articles, is available at https://go.pardot.com/e/272522/um-email-utm-source-newsletter/vhh4v/325179655?h=fYbpyFObU_WA2kqMpOQeUtkHkei29n6FWOY6dDuEt70)

A teachable moment doesn’t have to arise naturally; even a simulated one helped make British police officers 21% less likely to be hooked by a phishing email, in one of a series of behavioural nudges tested in 2018.

Nudges often only influence a small fraction of the target group because they preserve choice by definition, as opposed to incentives or penalties, but are still good value as they can be scaled up at very low cost.

Several trialled over the past year have produced much larger effects, however, according to the latest yearly research report from the Behavioural Insights Team, a company jointly owned by its staff, the British government and the innovation quango Nesta.

The most successful of several measures to raise cyber security awareness in the UK Metropolitan Police involved sending test subjects a fake phishing email.

Instead of malware, those who took the bait and entered their login credentials were directed to a set of simple tips on avoiding the real thing, an effort to stage a “teachable moment” when they were most receptive to information.

The training document was a set of simple cyber security rules of thumb, itself based on what the BIT has learned from past research about effective ways of getting written information to actually sink in.

Another test group was given the same document but without the set-up, and a third got information drawn from the official advice of the UK Centre for the Protection of National Infrastructure. The randomised trial involved over 17,000 officers and included a control group.

More emails “closely resembling actual phishing emails from past attacks” were sent again three weeks later, and then again after three months to see if the training stuck.

All three interventions had significant effects lingering three months after the training. Interestingly, the email containing official advice from the CPNI was second-best, reducing the number who fell for a mock phishing email by just under 17% compared to about 10% for the BIT-crafted email on its own.

Here in Australia, applications for rural teaching jobs tripled from 4.2% to 12.6% when a paper application form was replaced by a quicker and easier online version and combined with several other types of nudges known to be effective. The trial was run with the New South Wales Department of Education, three universities and the NSW Department of Premier and Cabinet’s own nudge unit, set up a few years ago with assistance from the BIT.

“To overcome some of the ‘friction’ in the application process, we made it easier to apply for rural placements by replacing the paper application with a partially pre-filled online form and increasing the number of rural schools that could be selected for students’ placements,” the report explains.

“We also tested a series of prompts, including: extra information about rural schools; encouragement to discuss the idea with family and friends; and a timely reminder before applications closed.”

In Moldova, another trial almost doubled the number of tuberculosis patients who kept up their treatment from 44% to 84% by giving them a “virtual version” of the recommended clinical supervision. This overcame the inconvenience of the normal method – doctor or nurse watching them take their medication every day.

Back in the UK, the BIT worked with the conservationist World Resources Institute, which wants everyone to eat less meat, and found food labelled “meat-free” was only half as popular as vegetarian options marketed using “more indulgent names” on the menu. The phrase “field-grown” was particularly appealing.

In another project, the UK’s international nudge company made a Bank of England report on inflation much more easily understood by readers. The best of four different versions they tested was a “relatable summary” using simple pictures and clear examples of what the inflation figures mean in everyday scenarios.

It led to a 40% boost to comprehension scores, compared to the standard BoE version, and a 16% increase compared to the central bank’s existing visual summaries.

A different project found repeat offenders in the UK are more open to voluntary support services designed to help them turn away from a life of crime, if the offer comes on their birthday – a bit over 4% compared to 2.6% who agreed when it came on a random day.

Why? “People’s hopes and motivations are higher around ‘fresh start’ events such as birthdays and the beginning of a new year, or even a new week,” according to the applied research company.

The report also includes information on commercialisation of BIT research and its involvement in the What Works Cities initiative in the United States, which encourages and supports experimental trials in local government: “In the first three years of the initiative, BIT has launched 99 trials in partnership with 37 cities. We believe this accounts for the majority of all field trials run by local governments in the US – ever.”

Other case studies involved encouraging over 11 million Indonesians to file their taxes early, netting over $2.6 million AUD; SMS reminders that nudge slightly more people charged with domestic violence in NSW to turn up to court; and an example where the BIT helped redesign traffic infringement notices for the one of the world’s leaders in issuing fines, the Singaporean government.

Nudging public servants

Peter Shergold, the former Australian Public Service chief and national president of the Institute of Public Administration Australia among several other roles, congratulated the BIT on “yet another impressive report” of a year’s work.

“It is increasingly obvious that the Team is developing a wealth of experience and evidence that will help governments to design and deliver public policy better,” Shergold said by email. “Public services around the world will benefit from the Team’s willingness to innovate and experiment.”

One interesting piece of recent work from the BIT is its Behavioural Government report, published in July in partnership with the Institute for Government, which looked at the impact of cognitive biases on decision-making in public service departments.

“We’ve since run training workshops with many UK government departments to help put these tools into practice,” says the 2018 round-up. “Preliminary results from that training suggest that it has reduced overconfidence among UK civil servants.”

The BIT also did some work on improving public service performance in October that was provided to the UK’s new Centre for Public Services Leadership. The yearly round-up explains:

“Our work suggests that interventions to develop leaders, such as development programmes, networking and feedback, can work. However, they must account for culture and context, offer practical insights and focus on behaviour change before, during and after any training or intervention.

“Real gains can be made by applying behavioural insights to how organisations and our public services work. Our work suggests careful nudges can spark inclusion, collaboration and innovation across organisations and help make us better leaders, managers and employees.”

British Minister for Implementation Oliver Dowd notes in a foreword that it is hard for ministers and public servants to say they aren’t sure of the best solution to a particular problem. “Yet sometimes, this is exactly what we should do,” he adds.

“That doesn’t mean doing nothing, or being frozen in uncertainty. Rather it means that governments, wherever possible, should test ideas empirically, including variations to see which version works best.”

Cabinet secretary and head of the UK civil service, Mark Sedwill, observes: “Sometimes the most serious barriers to more effective policy and practice are the presumptions in our own heads.”

Tags: behavioural economics, Behavioural Insights Team, cyber security, experimental, implementation, nudge, randomised controlled trial, RCT