This report has collated and synthesised guidance from the global behavioural science community on what has been effective so far in mainstreaming behavioural public policy. But these principles will need to be updated as policy challenges evolve, public administrations refine their operations, and behavioural science develops new insights and new methods. This section draws on active discussions in the field to suggest areas where we may see new good practices emerge in coming years.
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)
10. Emerging practices
Co-creating with citizens
Over time we can expect continued growth in citizens’ general familiarity with behavioural science – their “psychological capital” (Whitehead et al., 2014[1]). A more informed public opens up more opportunities for self-nudges: giving citizens the tools to shape their own choice environments (Reijula and Hetwig, 2020[2]). There may be opportunities for behavioural science experts to revise their role from being architects with an objective view of a policy issue, to being facilitators who actively convene coalitions of stakeholders to co‑create social change (John and Stoker, 2019[3]; Hallsworth, 2023[4]) – a shift in mindset that many designers and ‘design thinking’ practitioners have taken already (Mosely, Markauskaite and Wrigley, 2021[5]). This shift might require new skills and new operating models within government organisations (Einfeld and Blomkamp, 2021[6]), with behavioural science experts jointly participating in discovery, design, and evaluation activities alongside policy makers and stakeholders.
A more aware population might also become better at noticing when governments are using behavioural science interventions to influence their behaviour. How citizens interpret an intervention when they perceive one has been called ‘social sensemaking’ (Krijnen, Tannenbaum and Fox, 2017[7]). Recognising and understanding how citizens react to governments’ use of behavioural science is likely to become increasingly important for policy makers adopting behavioural science insights (Buttenheim, Moffitt and Beatty, 2023[8]). Minimising backlash and maintaining a generally supportive environment for behavioural public policy may require governments to be transparent in their use of behavioural science, and potentially to explore new mechanisms for accountability to the public.
Attending to diversity
Policy makers and behavioural science experts sometimes design solutions for what they deem to be an average or standard user or citizen. As the policy making community normalises the insight that people can be influenced by behavioural, cognitive, and social factors when they interact with government policies and programs, the next frontier for exploration becomes how those factors differ across groups in society – and how policy interventions should adapt and respond to those differences.
Behavioural science experts are increasingly expanding their evidence generation beyond ‘what works’ to ‘what works for whom and in what context’. Where data disaggregated by diversity characteristics is available, this expanded viewpoint can help policy makers understand if policies are reaching those most in need, and tackle complex social problems in a way that improves social equality (Buttenheim, Moffitt and Beatty, 2023[8]). As data collection expands, administrative datasets are linked, and analytical tools – such as machine learning algorithms – get more advanced, it will become easier for behavioural science experts to attend to diversity when diagnosing policy issues and measuring the effectiveness of possible solutions. Opportunities will also emerge for personalising interventions and changing how a policy is implemented based on features of the beneficiary (Mills, 2020[9]). Governments will need to ensure diversity data is collected, and then institute accountability mechanisms to ensure it is managed and used ethically and appropriately.
Policy makers are likely to be increasingly called upon to recognise their own position in the social systems they are trying to influence (Hallsworth, 2023[4]). In many cases, the behavioural public policy community does not represent the diversity of the community they are designing policies for (Liscow and Markovitz, 2022[10]). Recognising their own entanglement in a particular network of assumptions, norms, and values can help policy makers attend to diversity and make more effective policies. Governments could consider diverse hiring practices and participatory citizen engagements as mechanisms to help policy makers and behavioural science experts notice socio‑cultural differences and integrate those insights into their work (See Principle 12).
Tackling complex social problems
Early forms of behavioural public policy focused on insights and methods from behavioural economics and psychology, leading to a toolkit and approach that was very effective at optimising solutions to policy problems that were well-defined and relatively predictable. Over time, however, behavioural science experts have been called on to contribute perspectives and evidence to help governments tackle complex, long-term, multifaceted problems – such as entrenched disadvantage or greenhouse gas emissions – which do not lend themselves to the focused analysis and structured tools that sometimes characterised early behavioural insights units (Feitsma and Whitehead, 2019[11]; Straßheim, 2020[12]; Ball and Head, 2021[13]). Recognising this, behavioural science experts have increasingly expanded their mindsets and methods to embrace a broader sweep of behavioural, social, and cultural perspectives (Hallsworth, 2023[4]; WHO Regional Office for Europe, 2022[14]). Adopting ways of working first developed by ethnographers, designers, foresight practitioners, and others has the potential to help behavioural science experts recommend policy solutions that are resilient to changes in context, scale, and time (Schmidt and Stenger, 2021[15]). These broader approaches require broader skill sets, further expanding the list of skills needed to produce rigorous and useful behavioural science evidence.
Framing policy issues as the outputs of socio-technical systems is one fruitful lens that can complement a behavioural lens (Diaz Del Valle, Jang and Wendel, 2024[16]). Systems thinking offers policy makers tools and approaches to map, understand, and model complex policy problems by thinking through the actors involved, how they are connected, and the rules and norms that govern their interaction (Nguyen et al., 2023[17]; Chater and Loewenstein, 2023[18]). Systems thinking can help behavioural science experts identify leverage points where relatively minor contextual modifications might produce cascades of effects across a whole system (Meadows, 2008[19]). Being mindful of broader connections can also help behavioural science experts identify diverse solutions, monitor for unintended consequences, or spot situations where the effect of a successful behavioural change is counteracted by a compensatory effect elsewhere in the system. Participatory research methods may help behavioural science experts understand the scope and nature of the system they are trying to influence.
More testing before policy decisions
There is potential for new and emerging data collection and analysis methods to dramatically increase the amount and scope of testing that happens before decision makers settle on a preferred policy solution. Emerging methods in the behavioural sciences can enrich ex ante policy evaluation (such as regulatory impact assessments), helping policy makers understand the likely effectiveness of different policy options. A recent focus in behavioural science on replications and multi-site ‘megastudies’ – which test large numbers of interventions simultaneously using consistent methods (Duckworth and Milkman, 2022[20]) – will expand the knowledge base about what works in what context, giving policy makers greater confidence to trial an idea in their own jurisdiction. Greater confidence will also come from ongoing efforts to synthesise existing knowledge better, such as shared ontologies that enable artificial intelligence to accurately spot patterns across large corpuses of previous research (Michie et al., 2020[21]). Artificial intelligence tools will continue to mature, improving the depth and speed at which behavioural science experts can analyse big datasets to identify problems and suggest promising solutions. Knowledge brokerage and networking across policy makers and behavioural science experts – including those outside government – will be critical to realise the benefits of these methods.
More generally, there is scope for behavioural science experts to focus on fit-for-purpose methods: doing the most sensible research activity that meets policy makers’ immediate needs with evidence that is sufficiently rigorous (Varazzani et al., 2023[22]). This could involve testing using the full spectrum of methods available, or establishing enduring data collection and analysis structures that facilitate inquiries on short timeframes. Governments will need to adopt a behavioural lens early in the policy process, and across their operations, to ensure the structures are in place to produce this timely and useful evidence.
More adapting after policy decisions
There remain opportunities for governments to adopt more agile approaches to policy implementation. Even once a policy decision is made, there remains a role for behavioural science experts to support ongoing evidence generation efforts. Active monitoring of the behavioural outcomes of a policy helps policy makers understand if their expectations hold true once the policy is rolled out on the ground and at scale. A productive learning loop between staff implementing policies and staff advising on policies can ensure lessons are shared and acted upon in future policy cycles (Lowe et al., n.d.[23]). Policy makers balancing conflicting needs of evidence generation and ethical access to beneficial programs might benefit from newer approaches to impact measurement that integrate early feedback into the study design, enabling adaptation over time (such as ‘bandit’ protocols that move people into intervention groups that appear to be the most successful) (Hallsworth, 2023[4]). Government may become more open to adaptive learning if senior leaders actively support these approaches, if funding decisions encourage active monitoring and review, and if behavioural science experts are structurally integrated or richly connected with staff implementing and evaluating policies.
Looking at government operations, processes, and people
Behavioural science experts in government have mostly focused on achieving external behaviour change. They have turned relatively recently towards internal behaviour change: improving the operations, processes, and decisions made within government organisations (Grimmelikhuijsen et al., 2016[24]). In this space, as with all others, behavioural science experts are partnering with specialists who have focused on these issues for decades, such as human resources professionals and organisational psychologists. Governments that are open to experimenting with their ways of working stand to benefit from more efficient and effective processes that are people-centred and evidence-informed.
The interface between government operations and citizens is another fruitful area for further development. Adopting a behavioural lens to analyse citizen-facing services helps to identify unnecessary frictions that incur time, money, and psychological costs – on both citizens and the staff they interact with – and exacerbate social inequities. These ‘sludge audits’ (Sunstein, 2020[25]) offer a tangible, structured, and effective way to integrate insights from governments’ efforts to reduce administrative burden (OECD, 2009[26]) and design user-centred services (McGann, Blomkamp and Lewis, 2018[27]). This behavioural approach to analysing a process can also be adopted to look at internal operations, such as recruitment or procurement.
References
[13] Ball, S. and B. Head (2021), “Behavioural insights teams in practice: nudge missions and methods on trial”, Policy & Politics, Vol. 49/1, pp. 105-120, https://doi.org/10.1332/030557320x15840777045205.
[8] Buttenheim, A., R. Moffitt and A. Beatty (eds.) (2023), Behavioral Economics, National Academies Press, Washington, D.C., https://doi.org/10.17226/26874.
[18] Chater, N. and G. Loewenstein (2023), “Where next for behavioral public policy?”, Behavioral and Brain Sciences, Vol. 46, https://doi.org/10.1017/s0140525x23002091.
[16] Diaz Del Valle, E., C. Jang and S. Wendel (2024), Behavioral Systems, Busara.
[6] Einfeld, C. and E. Blomkamp (2021), “Nudge and co-design: complementary or contradictory approaches to policy innovation?”, Policy Studies, Vol. 43/5, pp. 901-919, https://doi.org/10.1080/01442872.2021.1879036.
[11] Feitsma, J. and M. Whitehead (2019), “Bounded interdisciplinarity: critical interdisciplinary perspectives on context and evidence in behavioural public policies”, Behavioural Public Policy, Vol. 6/3, pp. 358-384, https://doi.org/10.1017/bpp.2019.30.
[24] Grimmelikhuijsen, S. et al. (2016), “Behavioral Public Administration: Combining Insights from Public Administration and Psychology”, Public Administration Review, Vol. 77/1, pp. 45-56, https://doi.org/10.1111/puar.12609.
[4] Hallsworth, M. (2023), “A manifesto for applying behavioural science”, Nature Human Behaviour, Vol. 7/3, pp. 310-322, https://doi.org/10.1038/s41562-023-01555-3.
[3] John, P. and G. Stoker (2019), “Rethinking the role of experts and expertise in behavioural public policy”, Policy & Politics, Vol. 47/2, pp. 209-225, https://doi.org/10.1332/030557319x15526371698257.
[7] Krijnen, J., D. Tannenbaum and C. Fox (2017), “Choice architecture 2.0: Behavioral policy as an implicit social interaction”, Behavioral Science & Policy, Vol. 3/2, pp. i-18, https://doi.org/10.1353/bsp.2017.0010.
[10] Liscow, Z. and D. Markovitz (2022), “Democratizing Behavioral Economics”, Yale Journal on Regulation, Vol. 39, pp. 1217-1290, http://hdl.handle.net/20.500.13051/18278 (accessed on 29 September 2023).
[23] Lowe, T. et al. (n.d.), Human Learning Systems: A practical guide for the curious, Centre For Public Impact, https://www.centreforpublicimpact.org/partnering-for-learning/human-learning-systems/a-practical-guide-for-the-curious48hjg7 (accessed on 25 September 2023).
[27] McGann, M., E. Blomkamp and J. Lewis (2018), “The rise of public sector innovation labs: experiments in design thinking for policy”, Policy Sciences, Vol. 51/3, pp. 249-267, https://doi.org/10.1007/s11077-018-9315-7.
[19] Meadows, D. (2008), Thinking in Systems: A Primer, Chelsea Green Publishing, White River Junction, VT.
[21] Michie, S. et al. (2020), “The Human Behaviour-Change Project: An artificial intelligence system to answer questions about changing behaviour”, Wellcome Open Research, Vol. 5, p. 122, https://doi.org/10.12688/wellcomeopenres.15900.1.
[9] Mills, S. (2020), “Personalized nudging”, Behavioural Public Policy, Vol. 6/1, pp. 150-159, https://doi.org/10.1017/bpp.2020.7.
[5] Mosely, G., L. Markauskaite and C. Wrigley (2021), “Design facilitation: A critical review of conceptualisations and constructs”, Thinking Skills and Creativity, Vol. 42, p. 100962, https://doi.org/10.1016/j.tsc.2021.100962.
[20] Nelson, K. (ed.) (2022), “A guide to megastudies”, PNAS Nexus, Vol. 1/5, https://doi.org/10.1093/pnasnexus/pgac214.
[17] Nguyen, L. et al. (2023), “Implementation of Systems Thinking in Public Policy: A Systematic Review”, Systems, Vol. 11/2, p. 64, https://doi.org/10.3390/systems11020064.
[26] OECD (2009), Overcoming Barriers to Administrative Simplification Strategies: Guidance for Policy Makers, OECD Publishing, Paris, https://doi.org/10.1787/9789264059726-en.
[2] Reijula, S. and R. Hetwig (2020), “Self-nudging and the citizen choice architect”, Behavioural Public Policy, Vol. 6/1, pp. 119-149, https://doi.org/10.1017/bpp.2020.5.
[15] Schmidt, R. and K. Stenger (2021), “Behavioral brittleness: the case for strategic behavioral public policy”, Behavioural Public Policy, Vol. 8/2, pp. 212-237, https://doi.org/10.1017/bpp.2021.16.
[12] Straßheim, H. (2020), “The Rise and Spread of Behavioral Public Policy: An Opportunity for Critical Research and Self-Reflection”, International Review of Public Policy, Vol. 2/1, pp. 115-128, https://doi.org/10.4000/irpp.897.
[25] Sunstein, C. (2020), “Sludge Audits”, Behavioural Public Policy, Vol. 6/4, pp. 654-673, https://doi.org/10.1017/bpp.2019.32.
[22] Varazzani, C. et al. (2023), “Seven routes to experimentation in policymaking: A guide to applied behavioural science methods”, OECD Working Papers on Public Governance, No. 64, OECD Publishing, Paris, https://doi.org/10.1787/918b6a04-en.
[1] Whitehead, M. et al. (2014), Nudging All Over the World: Assessing the global impact of the behavioural sciences on public policy, Economic and Social Research Council, https://research.aber.ac.uk/cy/publications/nudging-all-over-the-world-assessing-the-global-impact-of-the-beh (accessed on 25 September 2023).
[14] WHO Regional Office for Europe (2022), Behavioural insights units. Setting up behavioural insights units for improved health outcomes: Considerations for national health authorities, World Health Organization, https://www.who.int/europe/publications/i/item/WHO-EURO-2022-4886-44649-63372 (accessed on 27 September 2023).