Consistent and evidence-based decision making by civil servants is the foundation of effective public administration. And yet civil servants, like all individuals, make decisions under conditions of limited time, information and resources. These constraints – combined with uncertainty, fiscal pressures, and increasingly complex policy challenges – shape how effectively governments deliver on their priorities. Behavioural public administration (BPA) applies empirical insights on human behaviour to improve how civil servants think, decide and act within these real‑world constraints.
This report presents a framework of behavioral practices that governments can adopt and adapt to their administrative contexts, to be more effective and deliver public value. BPA does not replace existing policy tools: it strengthens them by aligning policy, processes and systems with how people behave. Governments already apply many of these practices. Combined with a more explicit effort to adopt behavioural related tweaks or practices, they have the potential to deliver greater change. This was the case in two pilots conducted in the Italian Public Administration. These sought to apply a behavioural lens to improve effectiveness of existing practices, in particular:
Increase compliance with guidelines by prompting Italian public administrations to publish their organisational plans on a public database. This enhanced co-ordination between civil servants and helped them follow internal guidelines.
Set high-quality objectives aligned with public value. This pilot surveyed 45 386 Italian civil servants to understand how they set objectives and helped them set higher-quality objectives in their annual planning documents.
Throughout this report, BPA is identified as offering practical, low‑cost ways to improve how governments think, decide and deliver, particularly in the following areas:
1) Policy analysis and design: helping civil servants make better use of evidence and manage predictable cognitive pitfalls. Governments can:
Improve the use of framing when communicating risk-based outcomes. Decision-makers are prone to framing – they tend to be risk-averse when options are framed as gains, but risk-seeking when framed as losses, even when the underlying options are identical. Presenting options in neutral or combined gain–loss frames can support more consistent, evidence‑aligned decisions.
Standardise checklists as decision-support tools. Checklists support decision-making by helping civil servants’ focus attention on relevant factors and improve the transparency and consistency of reasoning.
Use citizen engagement mechanisms. Civil servants may assume that citizens share their preferences due to the illusion of similarity. Consulting citizens helps ensure policy overcomes this bias and reflects their lived experience. This practice reinforces the OECD’s Recommendation of the Council on Open Government to actively engage stakeholders in all phases of the policy-cycle and service design and delivery, and as operationalised OECD Guidelines for Citizen Participation Processes.
Support evidence-based policy making. Civil servants may show confirmation seeking – selectively considering evidence that supports their existing beliefs and overlooking contradictory evidence. Simplifying complex findings and introducing systematic “consider-the-opposite” checks helps counter confirmation bias and supports more rigorous policy development.
2) Policy implementation and project management, helping projects succeed. Governments can:
Strengthen project planning to mitigate common risks, such as overly optimistic timelines. Behavioural strategies like outside reviews, Red Teams, and pre-mortems combat these risks by promoting a realistic assessment of the time, cost and risks involved in a project.
Set smart anchors. Giving project managers realistic upper and lower‑bound estimates derived from past projects reduces biases, such as anchoring on overly optimistic initial assumptions.
3) Monitoring and evaluation, generating more policy-relevant evidence and identifying how behavioural factors influence implementation and outcomes in real-world settings. Governments can:
Embed behavioural research methods in monitoring and evaluation to identify behavioural factors and evaluate real-world impact of policies on citizen’s behaviour.
Institutionalise evidence-informed policymaking through training, mentoring, academic partnership, and accessible research resources.
Mainstream behavioural science with LOGIC: leadership, objectives, governance, integration, and capability building. These behavioural science principles ensure BPA is resilient over time.
4) People management can improve by understanding better what works and does not, to recruit the best person for the role and develop their capabilities once hired. Governments can:
Improve interviews by using standardised and structured interview questions with clear scoring criteria tied to expected behaviours in the role. This ensures consistent evaluation of candidates, less influenced by the interviewer’s biases.
Improve candidate assessment by reviewing written tasks through anonymous joint evaluation, randomising the order that responses are assessed, and allowing assessors to revisit their initial scoring. These practices minimise in-group favouritism, the halo effect, and ordering effects.
Encourage unsuccessful candidates to reapply. Rejections may disproportionately discourage candidates from under-represented groups. Encouraging candidates who are nearly successful to reapply in future increases the range and calibre of applicant pools for future roles.
5) Integrity in the public sector can be enhanced by applying behavioural science to incentives, decision points and office culture. BPA enhances standard legal, regulatory and institutional integrity tools, such as those articulated in the OECD Recommendation of the Council on Public Integrity. Governments can:
Promote ethical norms by communicating positive descriptive social norms when integrity is common or normalised; and communicating injunctive norms when it is not.
Encourage whistleblowing by creating trusted reporting channels with clear instructions. Whistleblowers may not report because of fear of uncertainty and catastrophising. Trusted and clear channels can help whistleblowers overcome these hesitations.
6) Processes within public administrations can be enhanced by removing sludge, the excessive and unjustified frictions that hinder a desired behaviour and cause inefficiencies. Governments can conduct sludge audits of internal administrative processes to find and remove sludge and friction.
Going forward, as civil servants increasingly rely on AI, with the appropriate safeguards, AI can further enhance the benefits of BPA, by making established practices of civil servants more systematic, consistent, transparent and traceable.
