This report’s good practice principles are relevant for all governments. They identify categories of activity that deserve attention regardless of policy topic, context, or the government’s maturity in mainstreaming behavioural science. How these principles should be implemented in practice will, however, be dramatically different in different governments, and in different organisations within the same government – and these practices will also need to change over time. Practices that may be critical for establishing a position for behavioural science in the policy system may become irrelevant, or even detrimental, as the approach becomes part of business-as-usual.
LOGIC: Good Practice Principles for Mainstreaming Behavioural Public Policy
![](/adobe/dynamicmedia/deliver/dm-aid--16962537-6cbe-435b-89b0-3a79cd6d4069/6cb52de2-en.jpg?preferwebp=true&quality=80)
9. Maturity journeys and case studies
Older teams tend to do more complex work
Governments and organisations are unlikely to mature at the same pace. We could assume, however, that those who began longer ago are more likely to be in a maturing phase. With that in mind, we can see a few differences in responses to the OECD’s surveys between newer and more established teams (see box below). In summary, older teams appear to apply more diverse skills and methods to a wider range of activities, drawing on richer external relationships and better data – but many still face challenges with translating their research and advice into policy impact.
Box 9.1. Differences in survey responses between newer and older teams
Teams emphasise slightly different drivers of success as they age (see Table 4.1). Reflecting their emerging status, newer teams were relatively more focused on highlighting the advantages of behavioural science. Older teams were relatively more concerned about recruiting and retaining talent.
More established teams more commonly reported doing almost all of the activities we asked about, suggesting it takes time for organisations to establish necessary processes, relationships, and tools (see Table 5.1). Respondents working in older teams were particularly likely to say they provided advice on policy design, suggesting they are working earlier in the policy cycle. Older teams were more likely to say they conducted their own experiments (which can require advanced resources, skills, and contacts) but also to say they were evaluating behavioural impact in other ways, suggesting a pluralistic and pragmatic approach to research methods.
The most established teams (those ten or more years old) were more likely than newer teams to report challenges with implementing results, scaling interventions, disseminating findings, and getting approvals for work (see Table 7.1). These respondents may have experienced more examples of their work not going as far as they had hoped, may be more critical of their own work, or may have higher expectations for what they want their work to achieve. The only difficulty they reported less than newer teams was getting access to broader outcome data, suggesting that it can be effective to mature data management practices over time.
When asked about skills in their team, older teams were more likely to report having every skill we asked about, suggesting a more advanced and technical team (see Figure 8.1).
Older teams were also more likely to report partnering with academics, consultants, and international organisations, suggesting that it takes time to build these external relationships (see Table 8.1).
Good practices at different stages
Different practices will help governments and organisations navigate the changing considerations and risks they are likely to face over time. The table below takes the five dimensions of LOGIC and suggests practices that may be more relevant in the emerging, growing, and maturing phases (see Figure 2.2). This is not intended to be prescriptive. Every government and organisation will have different starting points, policy systems, and end goals, meaning every path to mainstreaming behavioural public policy will be unique and context-dependent.
Table 9.1. Implementing the LOGIC principles across phases in the mainstreaming journey
Example practices that may be useful at different points in the process of mainstreaming behavioural public policy
Phase |
Leadership |
Objectives |
Governance |
Integration |
Capability |
---|---|---|---|---|---|
Emerging |
Launch a new effort with a visible statement from senior leaders. Present case studies to senior leaders to make the case for dedicated resources. |
Identify priority focus that lend themselves to a behavioural approach, considering the government’s agenda and what has been successful elsewhere. |
Allocate the mandate for driving the change management process to a person or team. Give a dedicated team of in‑house experts aou n initial funding window. Seek resources from external funding bodies. |
Audit what behavioural data is available on prioritised topics. Agree on ethical principles and draw on existing risk management protocols. |
Promote the value of a behavioural lens to build policy makers’ interest. Partner with external experts on projects. Focus the skills of an in‑house team on communications, policy making, knowledge brokerage, and change management. Create a network of behavioural science experts across the government. |
Growing |
Expand a coalition of champions throughout the policy system. |
Release a dedicated strategy for behavioural public policy that seeks both ‘quick wins’ and a longer-term vision. Expand the topics approached from a behavioural lens to include internal processes and external policies and services. |
Use accountability structures and funding arrangements to encourage policy makers to collaborate with behavioural science experts. |
Add simple checkboxes to consider behavioural science to policy processes. Consolidate ethical procedures that are appropriate for the context. Build bespoke data structures to build evidence on priority topics. |
Build policy makers' capability to identify when they need expert support. Encourage external experts to build the capability of government employees. Access a wider range of more advanced research skills. Expand networks of supporters across the policy community. |
Maturing |
Build behavioural science into regular briefings of incoming leaders. |
Integrate behavioural public policy into plans and strategies at all levels as part of business-as-usual. |
Revisit how behavioural science evidence is governed and funded to ensure coherence with other evidence generation practices and approaches. |
Build a people-centred, evidence-informed approach into the government’s expected practices, standards, and guidelines for policy making. Explore new ways to engage citizens in research methods and policy design. Build behavioural outcome measures into the government’s standard data collection activities. |
Optimise the balance of internal and external expertise through partnerships and networks. |
Case studies
Different paths to mainstreaming behavioural public policy are evident in the specific experiences of particular governments. The following sections – about Argentina, Australia, Canada, France, the Netherlands, New South Wales (Australia), and the United Kingdom – demonstrate various routes to success, with behavioural science experts and enthusiasts adopting practices and governance models that make sense in their context.
Argentina
In July 2021, with the support of the Presidency and the Economic and Social Council, a dedicated unit of behavioural science experts (Unidad Ciencias del Comportamiento y Políticas Públicas) was initially created in the Secretariat of Strategic Affairs, in the Cabinet Office of the President in “Casa Rosada”. After a cabinet reshuffle in August 2022, the unit was relocated into the Ministry of the Economy (Gobierno de Argentina, n.d.[1]) .
The mission of the unit is to promote the application of behavioural sciences for the improvement of public policies and government decision-making, in order to improve people's quality of life. Its objectives are: to explore and understand human behaviour and the way in which people make decisions to guide and design better policies; and to conduct dissemination, research, training, and education activities linked to behavioural science in the context of public policy.
As October 2023, the unit has executed more than 20 projects and interventions, and given more than 60 conferences, workshops and virtual and in presence interventions in national and international events.
The unit has selected projects and interventions considering the availability of data, the demand from policy makers, team members’ knowledge and experience, and the execution times that made the actions feasible. The unit has conducted projects across various policy topics, including developing a chatbot to promote COVID-19 vaccination (in collaboration with the United Kingdom’s Behavioural Insights Team), increasing digitising some interactions with the tax administration, stimulating breastfeeding in municipal health services, promoting detection and reporting of potential organ donors in hospitals, and other topics.
In addition to its own projects, during 2021 the unit organised a ‘Call for Projects on Behavioural Sciences Applied to Public Policies’, with the aim of promoting the design, implementation, and evaluation of public policies to solve management problems across all levels of government (national, provincial, and municipal). The aim was to promote collaboration among academia, consultancy sector and different levels of government institutions.
The unit created the Argentine Network of Behavioural Sciences, with the objective of providing a framework of excellence and support for the design, planning, and execution of the unit’s objectives and processes. The network brings together behavioural science academics and professionals, creates space for theoretical and methodological discussions, promotes training activities, enables exchanges with national and international experts, and coordinates with public officials and administrators. Approximately 200 professionals have joined and meetings have been held focusing on the unit's processes and areas of intervention. The network also aims to link its members with public administrators who wish to design public policies drawing on behavioural science evidence.
As a dedicated team in the centre of government, the unit has been limited in the number and scope of projects it has been able to execute in terms of the variety of application areas and population reached. Many topics of interest were discarded because organisational conditions did not support their realisation, and many others could not even be addressed or considered.
Some structural and institutional factors have hindered the unit’s work. Tackling these would take substantial and sustained effort on a broad political, legislative, and administrative scale in order to promote the use of behavioural insights as an additional tool for government regulation, which is usually focused on more traditional incentive- and information-based policies. Being located in a central agency of the federal government has made scaling difficult once an intervention has been tested and found to be effective. In many cases it has not been possible to transfer expertise to other government agencies to autonomously implement behavioural science tools.
The lack of reliable data, data structures, and processes for regular measurement and easy data sharing has complicated the design and testing of behavioural interventions. It has been difficult to complement or improve programs with behavioural interventions when these programs lack clearly defined objectives, sufficiently trained human resources, effective coordination across levels of government, and monitoring and evaluation plans. Staffing changes in policy areas due to cabinet restructuring have also caused delays and demanded repetitions during implementation.
Finally, the unit has experienced organisational cultures, mindsets, and attitudes that have complicated behavioural science activities. Some officials have resisted experimental evaluations, concerned about the appropriateness of denying a control group access to an intervention thought to be promising. A more general resistance to changes in policies and methods, and hierarchical leadership styles, have also introduced difficulties. Some senior leaders have seen behavioural science activities as a luxury or detail, amid other decisions or problems that are considered of greater urgency or magnitude. On occasion, government officials have been interested in behavioural science, but been overwhelmed by their workload and not felt management support to advance innovative strategies.
Nevertheless, the continuity of the implementation of behavioural insights into public policies and the support of international organisations may help to give momentum to behavioural sciences. The idea of the institutionalisation of a central, dedicated unit, linked with decentralised behavioural champions within different areas of the federal government, and partnering with state and local level authorities in the implementation of behaviourally informed policies, offers a scope for hope in the future development of behavioural sciences in Argentina.
Australia
Australia’s federal government has a long tradition of behavioural public policy. A key example is Australia’s compulsory superannuation scheme, which has counteracted present bias since 1992 by requiring employers to automatically put a proportion of employees’ wages into a personal retirement savings account (Kingston and Thorp, 2018[2]).
In the 2010s, inspired by the academic literature and the emergence of behavioural insights units around the world, various Australian government organisations began setting up their own in-house teams of behavioural science experts. In 2016, the government established the Behavioural Economics Team of the Australian Government (BETA), with the remit to apply and test insights from the behavioural sciences across the federal government. BETA is located within the government’s central policy coordination agency, the Department of the Prime Minister and Cabinet (PM&C), giving it the status and access necessary to start conversations across the policy system and convene coalitions of partners around priority policy issues.
Initially, BETA received seed funding and seconded staff from various government agencies, enabling it to demonstrate the feasibility of a central behavioural science unit. In 2017 a further time‑limited funding envelope of three years came via a broader government innovation program, the Australian Public Service Modernisation Fund. After a string of successful projects, and a 6:1 return on the government’s investment, the Prime Minister decided to maintain BETA’s staffing from 2020 on an ongoing basis within standard departmental funding.
BETA co‑develops projects with policy makers, scoping specific evidence production or brokerage activities that will meaningfully inform policy development, implementation, or evaluation. Under semi‑formal agreements, PM&C covers BETA’s staff time while the partner agency covers practical project costs, such as research participant recruitment or graphic design services. Splitting the costs in this way has “encouraged a stronger sense of engagement” from policy makers (Ball, Hiscox and Oliver, 2017[3]). These agreements also support BETA’s independence and integrity by distributing clear roles and responsibilities.
A clearly defined mission has helped guide BETA’s activities. The wording has evolved over time with changing priorities and contexts, and is currently “to improve the lives of Australians by generating and applying evidence from the behavioural and social sciences to find solutions to complex policy problems”.
BETA is a multi-disciplinary team of economists, psychologists, data analysts, policy experts, and project managers, with backgrounds in academia, the private sector, and government. The team’s diverse skill mix allows it to conduct and translate research that is timely and practical. With a staffing count ranging in recent years between 24 and 30, BETA has the critical mass to maintain its diverse skillset, take risks in project selection, and train new staff in its distinctive practices and methods.
BETA has a strong focus on conducting rigorous primary research. At the start, this was embodied in the role of Research Director, a prominent academic brought in to sit alongside the Managing Director (who had a more typical government background). This model later evolved to an Academic Advisory Panel of seven academics from various disciplines who advise on BETA’s projects and methods.
BETA is committed to transparency, including through a policy of publication by default for major research projects, presenting at public events, and maintaining a content-rich website. All of BETA’s quantitative trials and evaluations have been published on its website, with trial protocols and analysis plans pre‑registered. Most of BETA’s diagnostic and advisory work (based, for example, on qualitative methods or literature reviews) has remained within government to inform particular policy decisions. But all primary research strictly follows Australia’s National Statement on Ethical Conduct in Human Research, including independent review of research plans, methods, and materials by a university-based ethics committee. Mature and robust data management protocols and practices enable data sharing to occur with relative ease between government agencies.
Identifying the policy topics or processes that stand to benefit the most from a behavioural science perspective has always been a challenge. Secondments from policy agencies produced project ideas early in BETA’s existence. Later, BETA conducted ‘Opportunity Scans’ within government organisations, involving interviews with senior executives, surveys of staff, and workshops with in‑house behavioural science experts. Once BETA was more established, it triaged the requests it received using clearly defined criteria: the potential impact a project could have on Australians; how high a priority the policy topic was for the government; the fit between policy makers’ needs and the skills BETA could offer; and the resources available. Most recently, BETA has begun proactively pitching work in support of key government agendas.
Early on, BETA found that quick, small-scale trials helped to build capability, demonstrate value, secure further funding, and build trust (Ball, Hiscox and Oliver, 2017[3]). While these projects still have their place when engaging with new partners or leaders, BETA’s more sustainable funding and established reputation now enables it to also contribute a behavioural science perspective to conversations about complex, long‑term policy problems. Expanding beyond bounded problems to tackle more complex policy problems has entailed an expansion from focusing on policy implementation to include problem diagnosis and policy design, and from using quantitative to mixed methods.
Building the capability of the APS to deliver behavioural public policy has always been part of BETA’s mission. Early in BETA’s journey it developed introductory materials for a broad audience, including online courses and an interactive tool to analyse the behavioural drivers of a policy problem. More recently BETA has shifted focus to providing richer support for smaller groups of policy makers, including through a network of behavioural insights champions and a bespoke, one‑on‑one coaching service. Finally, project work is an opportunity to upskill partners in the uses and insights of behavioural science.
The ecosystem of other behavioural science experts and teams throughout Australian government agencies has also grown. These local experts have developed enduring relationships within their organisations and rich content knowledge in particular policy areas, leaving BETA to work on topics that are cross-cutting, particularly high priority for the government, or that do not have a dedicated team in their associated agency. BETA also provides a secretariat function for a cross‑government Behavioural Insights Practitioners Network, which periodically brings in-house behavioural science experts from 25 different agencies together to share lessons learned and best practices.
Canada
This section is adapted from a recent publication (Sanders, Bhanot and O’ Flaherty, 2023, pp. 9-23[4]).
The Government of Canada established its first behavioural science team in 2015 inside the Privy Council Office (PCO), a central agency responsible for policy planning. Like most new public sector behavioural science units at that time, the team was small at first and it acted primarily as a ‘knowledge resource’ on behavioural insights and experimentation. It wrote papers and gave many presentations about the promise of this new public policy tool for leaders across government.
The team's first evolution took place in 2016, shortly after Canada’s Prime Minister issued an Experimentation Directive for all federal department and agency heads (Impact Canada, 2017[5]). This directive asked Canada's most senior public servants to strategically invest departmental dollars in experimentation. It called for fostering “work environments … conducive to experimentation, innovation, and intelligent risk-taking,” and explicitly named the team at PCO as a source of support, noting it would be available to “help create the conditions for implementing rigorous experimentation approaches.”
In the ensuing weeks and months, myriad requests for support from across the federal government flowed in. Over the next two years (2017-2019), the PCO team initiated several partnerships with federal departments to design behaviourally-informed program improvements and test them with randomised trials, which demonstrated the real-world positive impacts of behavioural insights and methods.
The government’s use of behavioural science changed significantly in 2020. In March 2020, just days after Canada closed its borders in response to the rapid growth in confirmed COVID-19 cases in the United States and abroad, Canada’s Chief Public Health Officer, Dr Theresa Tam – a long‑time champion of social and behavioural science – publicly emphasised the need to employ insights and methods from the field to develop and implement the government’s response to COVID-19. Shortly thereafter, the Clerk of the Privy Council, the most senior public servant in the Government of Canada, reached out to the PCO behavioural science team directly for support on COVID-19 management and response.
The team was faced with a tremendous task: to produce rapid, evidence-based advice on issues that were evolving daily. Realising that randomised trials would take too long to produce relevant data, and that fieldwork could be risky given the rapid spread of COVID‑19 cases, the team turned to rapid surveying and other online data collection methods to produce high-quality evidence on a tight timeline.
As a first step, in April 2020, the team implemented a longitudinal survey series to develop a baseline understanding of Canadians’ thoughts, feelings, and behavioural responses to COVID-19, and to keep track of changes over time. This study – the COVID-19 Snapshot Monitoring Study (COSMO) – followed a cohort of 2,000 Canadians over 16 waves of data collection between April 2020 and December 2021. The first waves of COSMO were fielded just 3-4 weeks apart, generating new data and insights on an extremely rapid timescale. Eventually, COSMO was complemented by a series of ‘deep dive’ mixed‑methods online studies to explore barriers and drivers of specific protective health measures, and later by more traditional field research and experimentation efforts.
The team’s rapid delivery of valuable data and insights eventually drew attention from other areas of government. In 2021, Environment and Climate Change Canada (ECCC) and Natural Resources Canada (NRCan) reached out to explore applying the team’s research model to address climate change and improve environmental protection. In September 2021, ECCC, NRCan, and PCO together initiated the Program of Applied Research on Climate Action in Canada (PARCA), a program focused on understanding barriers to greater pro-climate and pro-environmental action in Canada, and producing evidence-based, behaviourally-informed strategies to address them. PARCA has now designed, implemented, and analysed data from more than 25 primary behavioural science research studies.
PCO’s team now supports six programs of applied behavioural science research using this new model, and has grown to approximately 40 researchers and policy analysts. The team is collecting data on how people across Canada think, feel, and respond to issues as diverse as anti-microbial resistance, immigration, and adoption of new digital technologies. The team is using these data to provide evidence-based advice for senior decision-makers and design improvements to existing programs and services, which can then be tested using rigorous methods, like randomised controlled trials.
The team has recently re‑initiated the Government of Canada’s Behavioural Science Community of Practice, a network of behavioural scientists working across Canada’s federal government. PCO co-chairs this network alongside the Public Health Agency of Canada’s Behavioural Science Office, which was founded in 2021. Based on a March 2023 survey of the network, more than 17 Government of Canada departments and agencies now have behavioural science teams or employ behavioural scientists, totalling more than 100 public servants working in behavioural science roles. While a handful of these behavioural science teams were established before March 2020, a majority have been established since that time.
Canada's behavioural science story demonstrates the importance of at least three factors in growing the practice of behavioural science within government and applying it successfully in the longer term:
Senior-level championship. The Government of Canada's first behavioural science team experienced important periods of growth and change following the Prime Minister's 2016 Experimentation Directive as well as Canada's Chief Public Health Officer's call to use social and behavioural science to address the pandemic in 2020.
Demonstrating success and value. If not for the successes of its early demonstration projects, the team may not have been called upon to support on COVID-19 response and management. Successful demonstrations of the value and impact of behavioural science were foundational to the team's eventual growth.
Flexibility and adaptability. The team was able to realise considerable value by pivoting from its more traditional methods (randomised controlled trials) to new, rigorous methods that could produce data and insights on important behavioural challenges on a rapid timeframe. Ultimately, this shift in the team’s approach resulted in a new way of working that has appealed to numerous partners across the federal government.
France
Following a series of exploratory projects that began in 2013, a team dedicated to behavioural sciences was established in 2017 with the creation of the Interministerial directorate for public transformation (DITP) within the French administration. This directorate is entrusted with the implementation of the government's public transformation program, which aims to foster a closer, simpler, and more efficient public action, ultimately leading to tangible improvements in the lives of French citizens and public sector employees. To achieve this, the DITP actively supports and ensures the delivery of high-impact policies, promotes the dissemination of innovative methods, and coordinates administrative action to simplify procedures and improve the quality of public services.
Comprising four behavioural scientists and one public policy expert, the behavioural sciences team receives public funding (through the Programme d’investissement d’avenir and then Fonds pour la transformation de l’action publique) primarily to assist administrations in optimising their policies. Since its creation, the team has supported around 50 projects across various administrations. These projects generally involve a diagnostic phase (literature reviews and ad hoc studies), a phase of intervention prototyping, and a phase of experimentation or evaluation under real conditions (such as a randomised controlled trial). Depending on the circumstances, this operating mode can be adjusted to address specific issues, such as during the COVID‑19 crisis. The team’s work involves external experts, primarily engaged through interministerial public tenders, allowing access to a broader range of expertise. Most of the projects are initiated at the request of ministries and administrations.
The work carried out during these projects is often complemented by contributions to specific subjects in collaboration with other expertise hubs of the DITP. This particularly involves addressing issues related to the simplification of administrative communications (such as forms and language) and human resources matters (such as organisational functioning and recruitment).
The implementation of these projects is also an opportunity to extend the use of behavioural sciences. This effort to educate and disseminate is embodied through various initiatives, including publications (experimentation reports and guides), training (teaching and conferences), and external communication. More broadly, this approach reflects a commitment to bringing together administrations and researchers through the promotion of evidence-based public policies.
Beyond these numerous initiatives and achievements, the main objective is to foster the internalisation of behavioural and more broadly scientific skills within administrations, ensuring that these skills are mobilised as early as possible in the design of public action.
The Netherlands
The focus on behavioural public policy in the Netherlands initially emerged as a bottom-up process. From 2004 to 2014, several ministries and other government organisations were already addressing the importance of behavioural public policy and, in some cases, applying behavioural science insights. The turning point was in 2014 when three advisory reports were released by different advisory councils, all urging the Dutch government to better use behavioural science evidence throughout the policy cycle. These reports were discussed in May 2014 during an interdepartmental strategic meeting with top level public servants from all ministries. The Behavioural Insights Network Netherlands (BIN NL) was established during this meeting with the task of promoting the exchange of knowledge and experiences among ministries and drafting a cabinet response to the three advisory reports. The main message of the cabinet response, sent to the Parliament in late 2014, was: “To make optimal use of behavioural science knowledge, it is important to systematically utilise this knowledge throughout the entire policy process, from policy development to policy implementation and supervision. Therefore, the ministries will invest in building (further) expertise in behavioural science knowledge.”
Since 2014, more ministries have invested in applying behavioural science insights. This resulted in a growing group of colleagues familiar with the basic principles of behavioural science, the development of various behavioural tools supporting behavioural analysis, over 100 tested behavioural interventions, and a multitude of behavioural advice for the development of regulations, legislation, subsidies, and implementation processes. Over the years, small behavioural teams (with two to four full-time equivalent staff) have emerged in many ministries and government services, and in some cases, medium-sized teams (five to 15 full-time equivalent staff), particularly in policy implementation and supervision. The Netherlands Scientific Council for Government Policy (WRR) advisory report “Knowing is not the same as doing” in 2017 (Keizer, Tiemeijer and Bovens, 2019[6]) emphasised the importance of considering people’s ‘ability to act’, which was subsequently included as a quality requirement in the Netherlands Regulatory Impact Assessment (IAK), addressing among other things default options, biases, and heuristics.
Since 2017, BIN NL has received funding from all 12 ministries, enabling the network to undertake various activities, including an annual congress (the ‘Day of Behaviour’), publications and an online community including a database of tested behavioural interventions. Every two years the network bundles the most recent interventions into a publication, “A Wealth of Behavioural Insights” (Netherlands Behavioural Insights Network, 2021[7]), which is sent to Parliament. The network’s role was expanded in response to the 2017 WRR report, encompassing not only facilitating the exchange of knowledge but also supporting and promoting the use of behavioural science evidence, including considerations for testing ‘the ability to act’.
While progress has been made, the normalisation of behavioural public policy within the Dutch government remains an ambition. Improving policy quality and leveraging behavioural science evidence go hand in hand. Through understanding the context in which people live and the factors that determine human behaviour, it becomes possible to create more effective policies. An important development in this regard is the implementation of the Policy Compass, which is the successor to the IAK. The Policy Compass is the central method to elevate the quality of the policy process within the Dutch government. The application of behavioural insights is integrally embedded in this Policy Compass. Additionally, the analysis and consideration of ‘the ability to act’ have been included as a quality requirement in the Policy Compass. A recent government-wide project has been initiated to stimulate the further development and integration of considering ‘the ability to act’ in the development of policies and legislation.
Finally, we are witnessing a more diverse application of behavioural public policy. In the early years, most work was done in the area of developing and testing interventions in policy implementation, such as trialling alternative emails and letters. Over the past few years, behavioural science knowledge has also been used to provide advice at the beginning of the policy process, when policy options are still in the development stage.
Leadership support has been crucial in reaching the current state, specifically from those directors that have been willing to employ behavioural experts, who bring new knowledge and are given the freedom to incorporate the behavioural perspective. Within this group of experts, ‘intrapreneurs’ – who work on establishing and expanding the behavioural function within the government – have been essential. Examples of intrapreneurs include the position of the BIN NL chairman, the recently created position of the government-wide coordinator on ‘the ability to act’, and several vocal coordinators of behavioural insights units.
The choice for a network structure to strengthen the decentralised teams has also been helpful. Most behavioural teams in the Netherlands are small. BIN NL enables them to join forces to organise activities with a government‑wide impact. This ensures that more policy makers will become aware of and learn about the added value of the behavioural science perspective. Working together to organise these activities also leads to an increase in the number of connections within the network. Consequently, it becomes easier for individuals to find relevant assistance and expertise, and to collaborate on behavioural projects.
While progress has been made, several challenges persist, with the three most significant being:
1. Further institutionalising the role of behavioural public policy advisors. Within many government organisations the implementation of behavioural public policy still relies on intrinsically motivated colleagues, without a formal position as a behavioural advisor. This brings the risk that knowledge and networks within these organisations may disappear when someone leaves, and will need to be rebuilt by a new colleague a few years later.
2. Integrating into the primary policy process. While the behavioural perspective is integrated via the quality requirement on ‘the ability to act’, time constraints often put pressure on the quick generation of solutions. The challenge is to take more time for thorough analysis or testing of policy variants for their impact.
3. Maturing of behavioural public policy. This involves various aspects, including the establishment of dedicated research facilities to enhance the quality and rapid execution of behavioural research.
New South Wales (Australia)
In Australia, the New South Wales government was an early adopter of behavioural public policy. In 2012 it invited the United Kingdom’s Behavioural Insights Team to help set up its own dedicated in‑house team: the NSW Behavioural Insights Unit (NSW BIU). Originally located centrally in the Department of Premier and Cabinet, the team later moved to the Department of Customer Service as part of a broader government strategy to focus on customer centricity.
The NSW BIU conducts its own projects focusing on direct, measurable improvements to government services and priority policy outcomes. The team complements these projects with an impact-oriented capability program named ‘Behaviour Smart’, which aims to build the capability and confidence of public sector organisations and public servants to use behavioural science in the delivery of government programs and policy. This capability program includes interactive workshops and training, regular drop‑in sessions named ‘BI clinics’ that solve specific policy or service challenges, project blogs and guides, and the flagship endeavours discussed below. The NSW BIU also supports the capability of public servants to reduce unnecessary frictions in government services with a sludge toolkit, sludge‑a‑thons, and a Sludge Academy.
Frontline service staff immersion in behavioural insights
Several public sector organisations in NSW have received assistance from the NSW BIU to introduce or deepen the application of behavioural insights within their organisation while solving service delivery challenges.
For example, NSW Legal Aid, a state-wide independent government agency that helps people in NSW with their legal problems, recognised the benefit that behaviourally informed service delivery could have. To support Legal Aid in a structured way, the NSW BIU invited four Legal Aid staff, ranging from policy officers to legal practitioners, to participate in a six-month immersion in the NSW BIU. Each staff member worked on a behavioural challenge relevant to Legal Aid, receiving direct support from NSW BIU staff to build their behavioural insights capability while making impact.
Each of the four immersion program participants designed and implemented an intervention using behavioural science, details of which were published on the NSW BIU’s blog (NSW Behavioural Insights Unit, 2023[8]).
As well as working on these four projects, Legal Aid NSW has established a Community of Practice, leveraging the expertise of these staff to identify other areas of application for behavioural science.
Impact focused learning: Behaviour Smart Bootcamp
The Behaviour Smart Bootcamp is a ‘learning by doing’ program. Through the Bootcamp, the NSW BIU provides 8 months of support to public sector teams, enabling them to both solve an immediate service challenge and build their own behavioural insights expertise so they can continue to apply the approach.
16 teams have completed the Bootcamp program in two cohorts. Teams have addressed diverse challenges, including minimising stonemasons’ exposure to silica dust, supporting farmers, and encouraging visitors to NSW to visit small businesses. Most teams graduate having implemented an intervention embedded in behavioural science and made concrete steps to evaluate it.
The Bootcamp provides teams with both theory and practice-based learning. It is delivered through:
Three in-person, interactive workshops. Here teams immediately apply what they are learning to their challenge, making progress on their project while also learning how to implement behavioural public policy.
A behavioural science expert mentor usually from the NSW BIU. Each team is assigned a mentor with expertise in behavioural science who provides direct support on the application of theory.
Self-guided resources. An online hub of resources is available for teams to access between workshops. These mirror and expand upon information provided in person, allowing teams to lift the capability of their broader agencies.
A regular community of practice. Bootcamp teams from all cohorts meet to learn about additional behavioural science methodology and share triumphs and tribulations in their Bootcamp projects.
Evaluating capability building and impact
The NSW BIU’s first priority in evaluating its Behaviour Smart capability focus is on identifying the impact of improved programs and services for people in NSW. However, the team has also assessed its impact on the skills and confidence of participating public servants.
In the 2023 Bootcamp the NSW BIU undertook a novel evaluation that included:
Identifying the specific skills and behaviours that a ‘Behaviour Smart’ public servant would apply or demonstrate (such as ‘using a framework to identify drivers and barriers to behaviour’ and ‘analysing behavioural data’).
Surveying Bootcamp participants before and after the program about how often they perform typical ‘behavioural insights practitioner’ behaviours in their work.
Creating a control group of similar public servants who were interested in behavioural science but did not participate in the Bootcamp. The control group were asked to complete both surveys, and a sample of them were matched to the treatment group of Bootcamp participants on previous behavioural insights experience and role seniority.
Using an anonymous but unique individual identifier in both surveys and matching as many participants in the pre- and post‑surveys as possible.
This evaluation allowed the NSW BIU to measure the unique impact of the Bootcamp on participants’ capability, by controlling for time and the impact of doing the survey itself.
In addition to improving services and having a direct impact on citizens, the BIU found from this evaluation that the Bootcamp improved capability. Bootcamp participants increased how frequently they used specific behavioural insights behaviours from before the program to after by 17.9% more than the control group. Confidence to use behavioural science at work also grew by 16.5% more in Bootcamp participants compared to the control group.
The NSW BIU is continuing to grow its Behaviour Smart program. This year the team will run another Bootcamp, iterating on previous programs to expand impact.
United Kingdom
In 2019, the Government Communication Service – the professional body for communication specialists across the United Kingdom (UK) government – created its own behavioural science team within the Cabinet Office to provide cross‑government support to major government communications campaigns.
When the team was initially created, it had three main areas of focus:
Cross-government consultancy offering. The team provided an in-house expert behavioural science consultancy service to departments across the UK government that were planning major communications campaigns.
Capability building and raising the standard of behavioural science across government. The team designed and delivered new training programs to upskill non‑behavioural scientists across government and local authorities in the use of behavioural science methods.
Crisis response. The team provided expert advice across government and within the Cabinet Office on crisis preparedness and response.
Following an internal restructure, the team paused its consultancy offering so that it could focus on campaigns run centrally from within the Cabinet Office.
Unlike many other teams across the UK government which specialise in the design and delivery of randomised control trials (RCTs), the Cabinet Office’s behavioural science team specialises in the use of secondary data, as well as behavioural science frameworks and theories, to make evidence‑based recommendations about the most appropriate course of action. This approach was found to be better suited to the fast-paced and sensitive nature of the work required at the centre of government, especially given that most departments already have very high‑quality RCT experts embedded within their own teams.
References
[3] Ball, S., M. Hiscox and T. Oliver (2017), “Starting a behavioural insights team: Three lessons from the Behavioural Economics Team of the Australian Government”, Journal of Behavioral Economics for Policy, Vol. Special is.
[1] Gobierno de Argentina (n.d.), Unidad de Ciencias del Comportamiento y Políticas Públicas, https://www.argentina.gob.ar/economia/planificacion-del-desarrollo-y-la-competitividad-federal/unidad-de-ciencias.
[5] Impact Canada (2017), Experimentation direction for Deputy Heads - December 2016, https://impact.canada.ca/en/reports/experimentation-direction-for-deputy-heads (accessed on 18 January 2024).
[6] Keizer, A., W. Tiemeijer and M. Bovens (2019), Why Knowing What To Do Is Not Enough, Springer Netherlands, Dordrecht, https://doi.org/10.1007/978-94-024-1725-8.
[2] Kingston, G. and S. Thorp (2018), “Superannuation in Australia: A Survey of the Literature”, Economic Record, Vol. 95/308, pp. 141-160, https://doi.org/10.1111/1475-4932.12443.
[7] Netherlands Behavioural Insights Network (2021), A Wealth of Behavioural Insights 2021, https://www.binnl.nl/home+-+en/knowledge/publications/bin+nl+publications/HandlerDownloadFiles.ashx?idnv=2239188 (accessed on 21 December 2023).
[8] NSW Behavioural Insights Unit (2023), Behavioural Insights in Practice: the Legal Aid NSW immersion program, https://www.nsw.gov.au/departments-and-agencies/behavioural-insights-unit/blog/behavioural-insights-practice-legal-aid-nsw-immersion-program (accessed on 1 February 2024).
[4] Sanders, M., S. Bhanot and S. O’ Flaherty (eds.) (2023), Behavioral Public Policy in a Global Context, Springer International Publishing, Cham, https://doi.org/10.1007/978-3-031-31509-1.