08 Sep 2017

Behavioural insights could play a key role throughout the entire policymaking cycle by helping governments nudge citizens towards certain actions. What are the key principles to apply to get it right?

Government agencies around the world face a common challenge: How can they design public policy interventions to achieve the desired results while delivering the best value for scarce taxpayer funds?

They have increasingly sought to tap into behavioural insights, relying on carefully controlled experiments and previous studies to understand how people respond to real-world stimuli.

Dr. Chen Kang, Visiting Professor at the Lee Kuan Yew School of Public Policy at the National University of Singapore, explained that human behaviour is not always predictable, as public policy has traditionally been informed by standard economic theories, which typically assumes that human behaviour is rational, individuals are unemotional and their preferences are stable.

However, he pointed out that human behaviour could often deviate from these standard assumptions due to psychological limitations. Incorporating behavioural considerations can help improve policy design and policy outcome.

Behavioural insights have been applied to a wide array of issues, according to Chen, including efforts to boost school enrolment of girls in rural India and encourage consumers to choose energy-efficient appliances in Singapore. Here are five key principles that policymakers should keep in mind when harnessing the power of behavioural science.

1. Leverage simple changes

Policymakers need to be open to the possibility that minor tweaks in policy design or engagement with citizens may lead to significant changes in behaviour.

Maya Shankar, a former senior White House advisor who chaired President Barack Obama's social and behavioural sciences team, recently described how swapping one word in an outbound email had a big impact on the number of veterans taking advantage of employment and educational benefits.

Instead of saying the veterans were eligible, the new email noted that they had earned the benefit. We found that the change of one word led to a nine per cent increase in access to the benefit, she said at an event hosted by the Council on Foreign Relations.

The example illustrates that increasing the take-up of a government scheme may not necessarily require big budgets for advertising or a fundamental rethink. In this case, the notion of having already earned a reward proved a more powerful motivator than the apparently less tangible idea of mere eligibility, a finding that allowed administrators to refine their message.

Another example cited in a report from the Behavioural Insights Team (BIT) was a trial conducted in 2015 by the Ministry of Manpower and the Central Provident Fund Board (CPFB) in Singapore. To encourage more people aged 55 to sign up for financial planning sessions, they improved on initial invitation letters by simplifying and personalising them. However, they also created a version with a pre-commitment message saying, We have reserved a place for you at the CPF Retirement Planning Service in . This variant increased attendance by 17 percentage points, while the one without the additional message only increased attendance by 5 percentage points. In light of the findings, CPFB decided to use the letter with the message and also launched a larger pilot study for added reassurance.

2. Apply throughout the policy cycle

Behavioural insights must be seen as more than a fad. It should not be a fashionable short-term foray by public bodies, said the Organisation for Economic Co-operation and Development (OECD) in its recent report Behavioural Insights and Public Policy: Lessons from Around the World.

However, the OECD observed that behavioural insights were still applied in a relatively late stage during the policy cycle, such as to fine-tune implementation when a policy is already in place. Behavioural insights can be applied to evaluate the effectiveness of implantation to feed lessons on what works and what does not into the early design of polices, it notes.

There is also a large opportunity to apply insights to change organisational behaviour inside and outside government. For instance, this could relate to ways of steering departments in applying cross-government agendas such as diversity, open government or good regulatory practice.

3. Carry out rigorous testing

Given that human behaviour is not always what we might expect, rigorous tests are essential for discovering the most effective approaches.

According to the OECD report, randomised controlled trials are one way to determine whether an intervention produces the desired change of behaviour. The trials should involve a sample size that is large enough and often include a standard or existing option to ensure that any apparent changes in behaviour are properly established. In explaining the need for this kind of rigour, the report said, the necessity of good or reliable data is key to applying behavioural insights.

For example, a study published by the World Bank in 2015 showed how a randomised controlled trial was designed to measure efforts to reduce household water consumption. The 5,626 households in Beln, Costa Rica were randomly divided into four groups, including a control group that did not receive any interventions. By doing so, researchers were able to see the effects of adding different kinds of low-cost stickers to water bills.

The study found that a sticker comparing a household's water use to the neighbourhood average was most effective, reducing consumption by between 3.7 and 5.6 per cent compared with the control group. Apart from confirming existing literature on the power of communicating local norms to encourage behaviour change, the study also produced a clear pathway to help the authorities stave off future water shortages.

4. Maintain transparency and address any ethical concerns

It is not hard to see why the idea of government bodies seeking to influence how people act can raise ethical concerns. It is therefore important that the public sector shows a high level of transparency.

The OECD advocated that government bodies should publish information about their work in journals or annual reports, including successful and unsuccessful trials.

Andreas T Schmidt, Assistant Professor at the Faculty of Philosophy in the University of Groningen in the Netherlands, explored such ethical dimensions in his paper The Power to Nudge, noting that some people were worried that furnishing the government with far-reaching powers to nudge its citizens leads to problematic power relations.

To ensure nudging is suitably transparent and democratically controlled, he suggested that the actions would have to be proposed and defended openly. In carrying these out, proper consultation and involvement of the public and relevant public interest groups would also have to be sought, while providing continued assessment of such policies and making them open to contestation.

5. Be aware of the limitations

When applying behavioural insights to policy challenges, practitioners should also be aware of constraints. Some applications could work for some but not all of the population. The results should also be monitored and evaluated to ensure apparent benefits are sustained, according to the OECD report.

BIT also highlighted a potential pitfall in its study EAST: Four simple ways to apply behavioural insights. Social norms can be used effectively to encourage people to pay their tax debts for instance, but "misapplication of behavioural insights might result in perverse effects". This occurs when policymakers unwittingly reinforce a negative behaviour by pointing out how widespread the problem has become.

Peter Ong, Head of Singapore's Civil Service, said: Ultimately, we are designing solutions for real people and not idealised actors, in a speech at the 2017 Behavioural Exchange Conference. Because human behaviour is so complex, we must use a combination of (behavioural insights) and different tools to understand and shape decision-making.

Putting principles into practice

Policymakers can improve their programmes or messaging by applying insights to how people will respond in different scenarios. With a clear emphasis on rigorous testing and transparency, public bodies should be looking for creative solutions early in the policy design process. They should also track results during implementation in order to achieve continuous improvement in policy design.

This piece was written by Daniel Hurst.