Applying Behavioural Science in the Italian Public Administration

6. BPA for monitoring and evaluation

Abstract

This chapter describes how behavioural science enhances policy monitoring and evaluation. Behavioural methods – experimental, quasi‑experimental and qualitative research – are essential to design policies and services, evaluate their impact, and learn why they succeed or fail. However, their impact is limited unless behavioural methods are mainstreamed through leadership, objectives, governance, integration and capacity building. Behaviourally-informed monitoring and evaluation enables evidence‑informed decision‑making across the policy cycle.

Key messages

• Embed behavioural research in policy. The subjects of policy decisions are people, and so it is important to study and monitor how a particular policy change would affect their behaviour. Understanding the choice architecture that people exist in, the barriers to desired behaviours, and the cultural contexts in which they operate are essential to crafting effective internal policy.

• Institutionalise evidence-informed policy to build civil servants’ capability, motivation and opportunity to do evidence-informed policy. Assess organisational capacities for evidence-informed policymaking, train civil servants through structured curricula, mentor new researchers, promote interaction between civil servants and academics, provide access to research through databases, research reviews, and informal knowledge sharing.

• Mainstream behavioural science to overcome barriers to adoption. Embed support for behavioural science in the leadership, objectives, governance structures, integration mechanisms and capability building of public administrations.

Why it matters

Public administrations monitor and evaluate policies and programs to track their progress, assess impact and recommend evidence-informed changes. For example, evaluation teams assess the best health communications to promote vaccination behaviour. Ministries of education assess whether new classroom techniques improve student behaviour. Regulators monitor the behaviour of regulated entities to understand barriers to effective compliance. This work promotes accountability, supports evidence-informed decision-making, and facilitates learning (OECD, 2025[1]).

Behavioural methods strengthen monitoring and evaluation by studying what works to change behaviour and why. Randomised controlled trials and quasi-experimental methods are invaluable for establishing causal relationships between policies, programs and outcomes. Qualitative methods reveal how and why policies produce the results that they do, uncover implementation barriers, and provide valuable context to refine future interventions.

However, behavioural research is under-utilised in public administration. Many civil servants lack the skills and authorising environment to apply behavioural research. Some public administrations deprioritise behavioural methods despite the value they bring. Public administrations can address these barriers by championing behavioural methods to evaluate impact, institutionalising behavioural science as part of the evaluator’s toolkit, and mainstreaming behavioural science across public administration.

Whom it involves

• Centres of government evaluation units who set standards, provide methodological guidance, and conduct or commission evaluations. These teams might be embedded in central co-ordinating ministries or agencies. Their work shapes the quality of research across public administrations.

• Policy teams in line ministries responsible for portfolio areas in their administration. These teams conduct exploratory research to understand policy problems and design options.

• Policy evaluation, analysis and research teams who specialise in evidence-informed policy and methods to synthesise research that informs policy design. These teams might conduct original research to inform policy, and articulate the uncertainties and gaps in existing research, or train policy teams in evidence-informed problem analysis and policy design.

• Independent oversight bodies who monitor and report on public administration performance, assess compliance and provide external scrutiny.

How to improve monitoring and evaluation

There is high-quality evidence that behavioural methods enhance monitoring and evaluation. Experimental and quasi-experimental methods are essential tools to assess policy impact. Qualitative behavioural research is valuable to understand why a policy or programme produced a certain outcome. Questions remain on how best to build upon these methods: such as how to take heterogeneity seriously, address gaps in the behavioural evidence-base, and assess hard to measure changes (Hallsworth, 2023[2]). Nonetheless, it's beyond doubt that behavioural methods bring immense value to public sector research and evaluation.

Evaluate the impact of interventions

Behaviourally informed government draws upon psychology and econometric methods to assess impact. Randomised controlled trials (RCTs) are particularly important because they provide a robust counterfactual to estimate an intervention’s impact. However, RCTs remain under-utilised. An OECD survey of 42 countries found that only 28% of those sampled used RCTs (OECD, 2020[3]). RCTs are of irreplaceable value to evidence-informed policymaking because of their strength at demonstrating impact.

However, recent work has recommended civil servants should not over-rely on RCTs because they may struggle to assess complex or constantly changing behaviours (Hallsworth, 2023[2]), and the statistical analysis of RCTs may be perceived as complex or beyond the skills of civil servants (Oliver et al., 2014[4]). This limitation in could be overcome with training and guidance (see box 6.1 below). As behavioural science units mature across governments, more units are sharing lessons learnt from conducting RCTs and supporting each other to deliver more RCTs.

Box 6.1. Case study: A guide to developing behavioural interventions for field trials in Australia

In 2016, the Behavioural Economics Team of the Australian Government (BETA) codified its experience conducting field trials in the Guide to Developing Behavioural Interventions for Randomised Controlled Trials. The guide supports civil servants to design and evaluate interventions by posing nine questions to ensure interventions are grounded in behavioural science, feasible, and testable. The questions, their purpose and corresponding phase are:

Discovery Phase

• What is the outcome of interest? To define the specific, measurable outcome of success.

• Can we accurately, directly measure the outcome using existing data? To determine whether reliable data is available to measure the targeted outcome.

• Can we deliver standardised interventions to a reasonably large, randomised population? To assess whether the intervention can be uniformly delivered at an appropriate scale for testing.

• Is an intervention in this space feasible? To evaluate whether practical, ethical, legal, political, and institutional factors support implementation.

Diagnosis Phase

• How can we (get out of the office to) better understand the behaviour? To gain insights from real-world users and contexts to understand behavioural drivers.

• Specifically, what behaviour is leading to the outcome? To identify the key actions or decisions that produce the outcome of interest.

• What is our theory, step-by-step, of the current behaviour? To construct a step-by-step explanation of why the current behaviour occurs.

• What interventions might influence the behaviour? To generate behaviourally informed ideas that could effectively shift the behaviour.

• What is our theory, step-by-step, of how and why that intervention will change the behaviour? To articulate the expected causal pathway from intervention to behavioural change and improved outcomes.

Source: (BETA, 2016[5]).

Where an RCT is not feasible, other methods may be more appropriate. Civil servants may employ so called quasi-experimental methods, such as a difference-in-differences, when they cannot randomise or need to monitor change over time. Where civil servants seek to understand why an intervention worked, or to make iterative improvements on their intervention, then more qualitative methods or process evaluations may be more appropriate.

The plethora of methods can be disorienting. Civil servants conducting evaluation often benefit from a framework to match their method with the behavioural problem they seek to solve. Initiatives such as the OECD’s 7 Routes to Experimentation (Varazzani et al., 2023[6]) and the LOGIC principles (OECD, 2024[7]) offer practical pathways for institutionalising this integration, helping civil servants select appropriate methods and build lasting capacities for evidence use.

Box 6.2. Primer: The Seven Routes Framework to choose the right research method

Seven Routes to Experimentation is a user-oriented framework by the OECD to guide civil servants and practitioners in selecting the most suitable methods for designing and evaluating public policy. This framework offers a structured way to align methodological choices with specific policy challenges, considering real-world constraints such as time, resources, and contextual relevance. Users approach the framework through the questions listed below to select an appropriate research method:

• Do you want to test if a solution works?

• Should the solution apply to global or local contexts?

• Do you want to compare changes across different groups?

• Do you want to measure changes over time?

• Do you want to test a causal relationship between a solution and its effects?

These questions form a map where each route represents one of seven experimental, quasi-experimental and observational methods:

• Experimental methods test cause-and-effect relationships. They test whether an intervention leads to a specific outcome. Examples of experimental methods include RCTs, A/B testing, Difference-In-Difference, and Before-After studies.

• Observational methods explore behaviours in context. They identify drivers and barriers and form the foundation for hypotheses. These methods are especially valuable when the aim is to understand behaviour without intervening directly. Examples of observational methods include longitudinal, correlational, and qualitative studies.

The framework does not prescribe a one-size-fits-all approach or a linear path through the policy cycle. Instead, it encourages methodological flexibility. Observational methods provide deep contextual understanding but cannot confirm causality. Experimental methods offer stronger causal inference but may be less suited for exploring the underlying reasons behind behaviour. Different methods can be used at different stages. Combining observation and experimentation often leads to more robust, actionable insights. The framework helps embed an evidence-informed and context-sensitive approach into behavioural policymaking. It empowers researchers to choose methods that are fit for purpose, grounded in behavioural science, and capable of producing meaningful, contextually relevant insights.

Source: (Varazzani et al., 2023[6]).

Civil servants often benefit from structured research training – both for exploratory research and evaluation. Guides and frameworks are useful; however, policy teams often need a critical mass of research expertise to effectively use these guides and implement their research methods. To be effective, capacity-building needs to go beyond individuals to encompass the structural, cultural, and leadership dimensions that shape how evidence is used within organisations.

Institutionalise evidence-informed policy

Fostering a research-based mindset across public administration is not simply a matter of making new research methods available. No method is self-executing. Public administrations may wish to build civil servants’ capability, motivation and opportunity to develop evidence-informed policy. A framework for this approach is the OECD Building Capacity for Evidence-Informed Policy-Making: Lessons from Country Experiences (OECD, 2020[8]) which shows how public administrations may build research capacity by:

• Diagnosing and evaluating organisational capacities for evidence-informed policymaking, which inform organisational strategies to build research capacity.

• Training civil servants to manage research projects by seconding research experts into policy teams and facilitating knowledge transfer by co-delivering projects; and by providing intensive skills training programmes which equip civil servants to interrogate and assess evidence.

• Mentoring civil servants to build their capacity to use evidence, promoting interaction between civil servants and researchers (e.g. mentoring programs, professional forums and seminars, lunch-and-learn sessions, and informal networks of civil servant researchers).

• Providing access to research through online databases, disseminating syntheses of policy research, commissioning research and reviews of research, organising seminars to present research findings, and providing access to knowledge brokers.

Barriers to evidence-informed policy often include the upfront investment in evaluation and learning, reputational risk from finding that policies may not work, poor data sharing practices, and inconsistent demand for evidence. In these settings, evidence may be perceived as politically sensitive rather than administratively beneficial. However, these challenges can be mitigated by mainstreaming behavioural science in government.

Mainstream behavioural science

Governments who embed behavioural science into administration can follow best practice outlined in LOGIC: Good Practice Principles for Mainstreaming Behavioural Public Policy (OECD, 2024[7]) which shows how behavioural perspectives can be systematically integrated across the policy cycle. These principles build the authorising environment for evidence-informed policy and build capacity to do research.

Table 6.1. LOGIC principles for mainstreaming behavioural public policy

Category	Principle
Leadership	Senior leaders actively promote and advocate for the integration of behavioural science in policymaking when appropriate. Managers cultivate and sustain senior leaders' support for applying behavioural science within their organisations.
Objectives	Senior leaders and managers collaboratively define the role of behavioural science in achieving the government's strategic goals. Managers systematically monitor the application and impact of behavioural science evidence on policies to facilitate continuous improvement. Senior leaders and managers promote the use of behavioural science to enhance internal organisational processes, rules, and incentives.
Governance	Senior leaders assign clear responsibilities for embedding behavioural science into policymaking and establish accountability mechanisms. Senior leaders and managers allocate sufficient resources to ensure that policy advice is informed by relevant and reliable behavioural science evidence.
Integration	Managers incorporate behavioural science into standard guidelines and procedures for policy development, implementation, and evaluation. Managers ensure that behavioural science is applied responsibly, transparently, and with high integrity to build and maintain trust among civil servants and citizens. Managers support the development of processes and structures for data collection and analysis that facilitate the diagnosis of behavioural issues and the evaluation of policy options.
Capabilities	Managers build civil servants' capability to apply behavioural science perspectives to their work. Managers develop sustainable mechanisms for civil servants to access behavioural science expertise. Managers ensure that behavioural science evidence can be useful to inform policy making processes through quality brokerage. Managers develop systems for disseminating knowledge and sharing best practices, such as creating networks of behavioural science experts and advocates.

Source: (OECD, 2024[7]).

Embedding behavioural science across the policy cycle enables evidence-informed policymaking. This includes ensuring access to high-quality data, building the expertise of civil servants to apply behavioural science and research, and fostering an environment that is open to evidence. High-quality evidence has limited value if it is overlooked or unavailable in critical moments (Banks, 2010[9]).

A truly evidence-informed approach to policymaking depends on a system that is not only open to evidence but structured to seek it actively. This requires institutional support and a policy process that integrates evidence at every stage of the policy cycle: from the initial identification of an issue, through the design of appropriate interventions, to the evaluation of outcomes (OECD, 2024[7]). A behaviourally informed administration embeds research and evidence-building in the policy process.

Behaviourally informed insights

• Embed research across the entire lifecycle. Monitoring and evaluation are one stage of the policy lifecycle, but civil servants can most usefully think about these activities at all stages of policy. Involve researchers and evaluators early and throughout in the process. Effective evaluations are embedded in policies and projects from the beginning, not as an afterthought.

• Institutionalise use of evidence, rather than relying on individual expertise only. Evidence-informed policymaking relies on organisational systems as much as individual skills. Institutionalising evidence use reduces reliance on individual expertise and contrasts status quo bias. Promote behavioural research into routines, governance structures and leadership expectations so that using evidence becomes the default strategy.

For further reading on how to improve monitoring and evaluation through evidence use, see for example: 7 Routes to Experimentation (Varazzani et al., 2023[6]); LOGIC: Good Practice Principles for Mainstreaming Behavioural Public Policy (OECD, 2024[7]); and Target, Explore, Solution, Trial, Scale (BIT, 2022[10]).

References

[9] Banks, G. (2010), Evidence-based policy making: What is it? How do we get it?.

[5] BETA (2016), Guide to developing behavioural interventions for randomised controlled trials.

[10] BIT (2022), Target, Explore, Solution, Trial, Scale. An introduction to running simple behavioural insights projects.

[2] Hallsworth, M. (2023), “A manifesto for applying behavioural science.”, Nature Human Behaviour, Vol. 7/3, pp. 310-322, https://www.nature.com/articles/s41562-023-01555-3.

[1] OECD (2025), Government at a Glance 2025, OECD Publishing, Paris, https://doi.org/10.1787/0efd0bcd-en.

[7] OECD (2024), LOGIC: Good Practice Principles for Mainstreaming Behavioural Public Policy, OECD Publishing, Paris, https://doi.org/10.1787/6cb52de2-en.

[8] OECD (2020), Building Capacity for Evidence-Informed Policy-Making: Lessons from Country Experiences, OECD Public Governance Reviews, OECD Publishing, Paris, https://doi.org/10.1787/86331250-en.

[3] OECD (2020), How Can Governments Leverage Policy Evaluation to Improve Evidence Informed Policy Making.

[4] Oliver, K. et al. (2014), “A systematic review of barriers to and facilitators of the use of evidence by policymakers”, BMC Health Services Research.

[6] 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.