You could say that the first one began in 2009, when the US government recruited Cass Sunstein to head The Office of Information and Regulatory Affairs (OIRA) to streamline regulations. In 2010, the UK established the first Behavioural Insights Unit (BIT) on a trial basis, under the Cabinet Office. Other countries followed suit, including the US, Australia, Canada, Netherlands, and Germany. Shortly after, countries such as India, Indonesia, Peru, Singapore, and many others started exploring the application of behavioral insights to their policies and programs. International institutions such as the World Bank, UN agencies, OECD, and EU have also established behavioral insights units to support their programs. And just this month, the Sustainable Energy Authority of Ireland launched its own Behavioral Economics Unit.
As eMBeD, the behavioral science unit at the World Bank, continues to support governments across the globe in the implementation of their units, here are some common questions we often get asked.
What are the models for a Behavioral Insights Unit in Government?
As of today, over a dozen countries have integrated behavioral insights with their operations. While there is not one model to prescribe, the setup varies from centralized or decentralized to networked. The UK (Prime Minister’s Office), Germany (office of Federal Chancellor in Policy Planning Unit) and the US (originally in the White House Office of National Science and Technology Policy) established first teams at the federal level, centralizing the provisioning of behavioral insights support at the executive level. The teams worked directly with implementing agencies on diagnosing, designing, testing, and evaluating behaviorally informed interventions. However, in all three examples, behavioral teams were also formed at the ministerial and/or state level. The federal teams provided overall strategic guidance, capacity building, and direct support to other government agencies.
The Netherlands adopted a ‘networked’ model, where ministries formed their own teams and the Ministry of Economic Affairs plays a secretariat role. Canada and Australia first established behavioral insights units at the state level before setting up teams at the sub-national levels.
In some countries, the units were first established at the ministerial level. One example is MineduLab in Peru, which was set up with eMBeD’s help. The unit works as an innovation lab, testing rigorous and leading research in education and behavioral science to address issues such as teacher absenteeism and motivation, parents’ engagement, and student performance.
Today, the UK has moved to a decentralized model, where ministries have set up their own teams, and the BIT became a limited company, jointly owned by the UK government, Nesta, and its employees.
The process tends to be fluid, starting first with a small team for a two to three-year trial period to demonstrate effect. Almost all units focused first on achieving quick wins to build buy-in, demonstrate effect, and increase awareness.
In short, centralization has the advantage of offering coordination and high-level support, but experience shows that some degree of decentralization is inevitable, given that most behavioral interventions are conducted at the implementing agency level.
What should be the structure of the team?
Most units start with two to four full-time staff. Profiles include policy advisors, social psychologists, experimental economists, and behavioral scientists. Experience in the public sector is essential to navigate the government and build support. It is also important to have staff familiar with designing and running experiments. Other important skills include psychology, social psychology, anthropology, design thinking, and marketing. While these skills are not always readily available in the public sector, it is important to note that all behavioral insights units partnered with academics and experts in the field.
The U.S. team, originally called the Social and Behavioral Sciences Team, is staffed mostly by seconded academic faculty, researchers, and other departmental staff. MineduLab in Peru partnered with leading experts, including the Abdul Latif Jameel Poverty Action Lab (J-PAL), Fortalecimiento de la Gestión de la Educación (FORGE), Innovations for Poverty Action (IPA), and the World Bank.
Most units have an advisory body consisting of government officials, academics, and experts to provide direction and support to the team. As the field is still evolving, leveraging external expertise through partnerships can bring the latest thinking, experience, and knowledge in house.
How to start?
Enlist a champion: Having a high-level official within the government helps drive the agenda, bring other stakeholders together, and navigate the authorizing environment efficiently.
Secure a two to three-year commitment. The start-up phase tends to be slow, especially if the team is sitting outside implementation agencies. Enough time needs to be factored in to assemble the team, establish an advisory committee (inter-ministerial, academia, research think tanks, private sector), build internal awareness, identify the 3-4 policy areas (which can be priority- or demand-based), conduct diagnostics, identify data sources needed to measure and evaluate, and design and test the interventions.
Go for low hanging fruits. It is often useful to select policy issues that can be tested easily, quickly, and at low cost. It is imperative to be able to demonstrate effectiveness within the first year. Identifying programs that have data available (i.e., administrative, household), an established communication infrastructure with beneficiaries (SMS, mail, email, centers), and enthusiastic implementers will allow the team to move quickly in diagnosing, testing and implementing the intervention. Furthermore, interventions that improve effectiveness of the service delivery tend to resonate with the mission of the implementing agencies.
Adapt behavioral insights to strengthen public trust: Nudging in public policy continues to face concerns related to manipulation, paternalism, and the removal of choice. However, as Sunstein and other proponents argued, all government programs are at some level applying choice architecture and nudging. Setting defaults and program design choices more deliberately, guided by science, can improve service delivery and increase the well-being of citizens. But it is important to appreciate that critics remain. Therefore, providing transparency and accountability for behavioral interventions is essential to sustaining or building trust in government. Nudges and other interventions inspired by the behavioral sciences need to be publicly disclosed and debated, as is the case for public policies more generally. Research by George Loewenstein and colleagues shows that disclosing “nudges” does not make them less effective.
Just as the global recognition of behavioral science opened the door for systematically applying and packaging insights from psychology, social psychology, neuroscience and other fields; institutionalizing behavioral science in governments could ensure systematic, scalable, and sustainable use of these insights in a transformative way. These are exciting times for behavioral science practitioners and enthusiasts; the field is growing fast and the practices of collaboration and knowledge-sharing are causing a ripple effect in all areas of public policy around the world. We no longer are asked whether governments should have behavioral insights units, but rather when and how.
Enroll here to get fresh Behavioral Science news from the World Bank