Harnessing habits: behavioural economics' impact from commuting to court dates

By David Donaldson

May 30, 2018

Recent experiments by the NSW Behavioural Insights Unit show governments how they can harness the increasing body of knowledge around how we respond to different interventions.

It might seem surprising that among all the administrative process and cost of the justice system, something as simple as a well-timed SMS reminder could help the courts work better.

But it can, according to the NSW Behavioural Insights Unit.

Texting defendants in family violence cases before their scheduled appearance reduced the number of no-shows by almost one quarter, reveals the BIU in a report detailing some of the findings of its experiments over the past two years.

Whereas 17.6% of defendants in the business as usual group failed to turn up to their court date, that number fell to 13.5% among those who received an SMS reminder.

This led to cases being finalised faster than they would otherwise have been, reducing case duration from 74 to 69 days on average. This frees up court and police resources.

Those expecting a deeper impact will be disappointed, however. On the most important measure, the results showed little change. There was a small reduction in the number of participants who committed another domestic violence charge within six months of the trial, and a small reduction in the number of breaches of apprehended domestic violence intervention orders, though neither was statistically significant.

Commuting outside peak hour

One of the BIU’s experiments set out to kill two birds with one stone — encouraging Department of Premier and Cabinet staff to take public transport outside peak hour, reducing crowding on trains and buses while encouraging more flexible work practices.

And it was successful — morning arrivals outside peak times increased by 8.6 percentage points.

Overall, the trial resulted in 550 additional instances per month where DPC team members avoided commuting to and from work during peak hours. Importantly, six months after the trial ended, staff had continued in their flexible working patterns.

A few interventions helped reach this result.

Normally, Outlook calendars show staff as available from 8am to 5pm. The researchers adjusted the default settings for ‘active hours’ on all staff Outlook calendars to reflect core hours of 9.30am to 3.30pm.

Many workplaces struggle to embed flexible work in their organisational culture, with employees often thinking they will be seen negatively for taking advantage of rules allowing flexibility. At DPC, data showed staff work hours most closely matched those of their directors, so the BIU and DPC Human Resources showed the directors this data, and prompted them to speak to their teams about flexible work.

The BIU also ran a team-based competition in the workplace to encourage staff to avoid peak-hour travel. It incorporated several behavioural insights elements, including social comparison, social norms, and incentives through competition points and a competition prize. There was performance feedback through a competition leaderboard that showed each team’s ranking in comparison to other DPC teams.

The results were reportedly more noticeable among women. Although both women and men were responsive to the interventions, the impact was particularly pronounced with female staff. It is likely the competition had a legitimising effect on women’s preferences for flexible work behaviour, the researchers believe.

“A key innovation was to demonstrate that behavioural insights can be applied to shift strongly resilient and habitual commuter behaviour, via the indirect mechanism of workplace levers,” the report argues.

“This is a world-first.”

Rural teaching placements

Attracting teachers to work outside the city can be a challenge for education departments, so the BIU set out to increase the number of students applying for placements during training.

A combination of changes — simplifying the application process and sending students behaviourally-designed information packs and reminders prior to the application deadline — tripled the number of students applying.

Making other options easier and increasing the information available is important in shifting people away from status quo options, say the researchers.

Another trial will test whether trainee teachers are more likely to apply for a rural placement if the opportunity includes the option to go with a peer or buddy.

Fines for littering

Reducing litter by 40% by 2020 is one of the NSW Premier’s Priorities, yet people are less likely to pay fines for littering out of their car on time than other traffic-related fines such as speeding or parking.

But something as simple as tweaking the fine notice can have an impact on the number of people paying on time and avoiding additional fees. This reduces administration costs for government.

Improved layout and the addition of messaging about the environmental impact of littering on the fine notice sent out to litterers led to an increase of five percentage points in payments made on time. Other approaches, such as a letter highlighting the personal consequences of being reported for littering, also led to an increase, though were not as effective.

If everyone in the trial period had been sent the best performing penalty notice, there would have been 820 more payments within three months and 777 fewer recipients receiving additional fines. The NSW Environmental Protection Authority now plans to use it as the default notice.

Class attendance

Every year a significant number of apprentices fail to complete their course, so the government is keen to increase the number of classes attended.

For many, there’s a disconnect between what they learn in class and what happens at work, so the BIU tested sending weekly updates to their employers on what they are learning in class.

The trial increased class attendance by three percentage points — though this had no effect on the number of students dropping out.

Null results

Null results — where the experiment was found to have had no effect — aren’t often reported because they’re not as exciting, but they’re important too. After all, knowing that something doesn’t work can be just as useful as knowing what does work.

The BIU wanted to encourage participation in planning processes, especially among groups who don’t normally get involved. But they found that redesigning the planning notice sent out to residents to make it more easily readable and include artist’s impressions of the new zoning had no statistically significant effect.

Another experiment tried encouraging kids to attend more sessions of a healthy living program aimed at reducing childhood obesity. The program works, so the researchers thought that increasing attendance might lead to stronger results. Text message reminders, a prize draw and simplified goal-setting did increase participation, but this did not translate into improvements in health or self-esteem.

About the author
Inline Feedbacks
View all comments

The essential resource for effective
public sector professionals