Applying Behavioural Science in the Italian Public Administration

Annex A. Practices for behavioural public administration

 

Theme	Practice	Rationale
Analysis and design
Present policy risks mindfully	Present both loss & gain frames simultaneously, e.g. This policy will save 100 lives (preventing 100 deaths). Codify this in briefing writing guidelines.	Civil servants may display loss aversion, over-weighting the value of safe, incremental changes over riskier options with greater expected value.
	Where joint and neutral framing is not possible, Frame risk-based outcomes according to loss/gain frame deliberately.	Civil servants may be susceptible to framing – more risk averse when options are presented in a gain frame, and more risk seeking when presented in a loss frame.
Reduce noise in decisions	Implement decision-hygiene protocols, such as case conferences where multiple reviewers first decide independently and then develop a consensus, relative judgements that assess options relative to each other, and algorithms that support, but never replace, human decision making.	Civil servants may display decision-making noise – random, unwarranted variation in choices that are expected to be uniform and consistent. Excessive noise can make these decisions unpredictable, unfair and inaccurate.
Introduce structured project decision points	Develop checklists that guide public administrations to consider all relevant factors and to assess the quality of their reasoning and decision-making process, to be evaluated and refined moving forward.	Checklists help contrast cognitive biases and decision-making errors in public administration by ensuring that key procedural steps are systematically addressed. They also promote reflective and transparent reasoning.
Strengthen diversity of perspectives	Engage with citizens directly and early in policy design to understand what people think, feel and decide.	The illusion of similarity may lead civil servants to assume that citizens share their own preferences and circumstances, misunderstanding how citizens will respond to their policies and programmes when they are implemented.
	Harness the methods of citizen participation to incorporate their unique perspectives and outside view (see above). This helps challenge the assumptions of civil servants whose mental model may capture only some facets of the policy problem.	The curse of knowledge may make it difficult for civil servants to understand what it’s like to not know something once you know it. Civil servants may struggle to imagine what it’s like to not share their policy expertise, leading them to develop policies, services and communications that are hard to engage with.
Improve interpretation of evidence	Prompt civil servants to Consider the Opposite (a structured challenge to their assumptions) by asking “What are some reasons that our initial judgment might be wrong?” and providing the best available evidence why.	Confirmation seeking may cause policy analysts to selectively gather and interpret evidence that supports their existing beliefs. Analysts may overlook contradictory evidence and emphasise evidence that agrees with their decision.
	Ask what scenarios could prove a judgment wrong.
	Simplify complex evidence, while still providing details, to reduce the scope for confirmation seeking to bias reader’s interpretation of results.
Reduce messenger-driven bias	Anonymise and remove or separate authorial markers, such as names and branding, from documents where their identity is not relevant.	The messenger effect may lead readers to assess evidence-informed on irrelevant attributes of the messenger rather than the substance of the evidence.
Use behavioural science for policy design	Apply behavioural frameworks such as ABCD, COM-B, or EAST to make policy more effective.	Behavioural frameworks enable evidence-based policies that account for how people make decisions in the real world rather than assuming totally rational behaviour.
Implementation
Strengthen project planning processes	Invite outside review of project plans from other teams and public administrations, this provides an independent perspective less prone to optimism and planning fallacy.	Civil servants may show optimism bias by systematically underestimating their project’s required time and costs while overestimating its benefits, focusing on the best-case scenario at the expense of more realistic, sometimes pessimistic, scenarios. Civil servants may show planning fallacy, predicting that their projects will be implemented as planned, even when similar projects have, historically, not gone according to plan.
	Organise “Red Teams” who act as adversarial collaborators, responsible for finding flaws in a plan to make it more effective and resilient. Effective Red Teams are given scaffolding, and their outputs are adjudicated. So, appoint a referee to ensure the Red Team’s criticisms are relevant and to hold the policy team accountable for actioning this feedback.
	Conduct pre-mortems that legitimise and encourage worst-case scenario planning within project teams.
Improve calibration & accuracy of estimates	Set smart anchors that reflect the best reference-class forecast available; or reflects a pessimistic anchor to mitigate unrealistic optimism; or provide two estimates to reinforce the uncertainty of the decision.	Civil servants may be anchored by initial estimates, failing to update their later assessment away from the initial anchor.
	Estimate project time, cost and outcomes using a reference class forecast which is based on the historical performance of similar projects.	Civil servants may make mis-calibrated estimates that are systematically under- or over-confident, leading to inaccurate forecasts, poor planning and difficulty learning from past outcomes.
	Encourage civil servants to make concrete estimates, e.g. that a policy has a 70% chance of succeeding, rather than vague estimates like “a good chance of success”.
	Make estimates within a range, rather than a single point, to reinforce the uncertainty of the forecast.
	Systematically gather civil servants’ forecasts and provide feedback on their estimates to improve their calibration.
	When multiple civil servants make independent estimates of an outcome, harness the wisdom of crowds by taking the average of multiple estimates.
Mitigate sunk cost thinking	Build explicit breaks into a project with go/no-go decisions at major milestones that only consider future costs and benefits, and which rule out consideration of past costs.	Public administrations may display sunk cost thinking to continue ineffective programmes to “justify” prior investments even though these costs are already spent and there’s no recovering them.
Monitoring and evaluation
Embed behavioural research in policy	Use behavioural research methods, such as RCTs, to identify relevant behavioural factors and evaluate real-world impact of policies.	Behavioural research methods provide robust evidence on what works, thus helping public administrations design more effective, evidence-informed, and context-sensitive policies.
Institutionalise evidence-informed policymaking	Encourage systems that enhance civil servants’ capability, motivation, and opportunity to apply evidence (e.g., through training, mentoring, academic partnership, accessible research resources)	Strengthening institutional support can ensure that evidence use becomes a routine in policymaking processes, not relying on individual initiative or temporary and external expertise.
Mainstream behavioural science within public administration	Integrate behavioural science principles into leadership development, governance structures, and organisational objectives to promote sustained adoption.	Embedding behavioural approaches in core administrative processes can help overcome implementation barriers and ensure long-term commitment to behaviourally informed policymaking.
People
Make flexible working the default in job ads	Prompt hiring managers to advertise flexible working options, attracting a larger number of candidates, especially more diverse candidates to public sector knowledge work.	Many hiring managers may consider onsite work the default for all new positions.
Conduct high-quality interviews	Conduct interviews with a panel of interviewers, avoid single-interviewer assessments	Structured, panel-based interviews can reduce individual bias and increase fairness and reliability in candidate evaluations. Standardized questions can improve consistency and predictive accuracy.
	Use standardised, structured interviews questions and scoring criteria.
Improve the assessment of written application tasks	Use joint evaluation and anonymisation to review all candidates’ responses to written tasks per question, rather than per candidate.	Various biases can deter civil servants from hiring the best candidate for the role: In-group favouritism may lead hiring managers to prefer candidates who share superficial characteristics with themselves. Halo effect may prompt civil servants to assess a candidate based on a single salient attribute or response to a question rather than a wholistic of all their relevant capabilities. Ordering effects may cause civil servants to assess candidates differently depending on whether they assess the candidate at the start, middle or end of a process.
	Randomise the order in which assessors review the candidates within each written task.
	Allow assessors to revise their earlier assessments of written tasks, once they’ve reviewed multiple candidates and developed a sense for the quality of all the responses.
Foster a diverse pool applicant to recruit the best candidate	Encourage candidates to do their best throughout the application. Send standardised messages to all candidates which can have a disproportionately positive effect for under-represented candidates.	Applicants may be short-term focussed, over-weighting costs of applying for or re-applying for a role at the expense of the longer-term benefits of potentially securing the role. Motivational messages can help overcome that barrier.
	Encourage all unsuccessful candidates to reapply if they reached a late stage of the application. This can have a disproportionately positive effect for under-represented candidates who may be less likely to apply gain.	Applicants may underestimate their suitability after an initial rejection. Encouraging all unsuccessful candidate that reached a late stage of the application to reapply and help contrasting self-selection bias in future application.
Integrity
Promote ethical norms	Communicate expectations that make clear what behaviours are expected and desirable.	Descriptive social norms communicate to civil servants what behaviours are common or normalized.
		Injunctive social norms communicate to civil servants what behaviours are acceptable or laudable.
	Signal that undesirable behaviours are not tolerated.	Descriptive and injunctive social norms exert a significant impact on unethical behaviour, which may spread if it appears to be tolerated.
Improve measurement of sensitive behaviour	Use anonymisation and indirect questioning to reduce social desirability bias, collect better data and develop a more accurate understanding of corruption.	Social desirability and the need to present themselves favourably may lead civil servants to under-report sensitive issues.
Strengthen whistleblowing systems	Introduce trusted and confidential reporting channels, such as a dedicated contact-point, and ensure visible follow-up.	Unclear procedures, fear of retaliation, and doubts about consequences are major barriers. Trusted, confidential channels can help employees to report wrongdoing.
Leverage leadership at all levels	Train managers to act as integrity leaders, embed integrity into leadership competencies, and encourage open conversations about ethics.	Leaders are powerful role models. When they visibly prioritise integrity, employees are more likely to adopt and sustain integrity norms.
Processes
Identify and reduce excessive burdens	Conduct sludge audits of internal public administration processes to measure, prioritise and remove sludge and friction.	Sludge – unjustified procedural requirements – can create administrative burdens that make processes unreasonably onerous to comply with.
Simplify and streamline processes	Embed UX/CX experts in business improvement teams to remove unnecessary frictions from internal public administration templates, forms and processes.	Civil servants may underestimate how friction, minor inconveniences and pain points can add up to make a service disproportionately unfriendly to users.
Improve clarity and accessibility of communication	Shorten and simplify communications and requests between public administrations.	Civil servants may underestimate how effective it is to make short, simple and clear requests of other civil servants. Long, legalistic requests run the risk of being ignored or skimmed over.
	Make the key ask of a communication or request prominent.
	Use formality effect (best-practice communications are often simple and formal).