Applying Behavioural Science in the Italian Public Administration

3. BPA to set better organisational objectives: a pilot from Italy

Abstract

This chapter describes how behavioural science can improve the quality of public administration’s objective‑setting. It describes a large‑scale behavioural pilot in Italy that help public administrations short-list objectives for their annual planning documents. The pilot found that clear, concrete feedback with examples were generally more effective at setting high-quality objectives aligned with public value. Guidance based on exhortations from public administration leadership to set high-quality objectives were less effective. The pilot suggests that behaviourally informed guidance can address civil servant’s social, informational and institutional barriers to setting better objectives.

Key messages

• Recognise the setting of high-quality administrative objectives is not straightforward. Civil servants and public administrations operate in contexts of shared responsibility and often on timescales that do not fall neatly into review cycles. In this context, broad and non-specific directives are likely to result in misinterpretation or poorly defined objectives intended to fulfil the process. Guidance should help civil servants focus their attention on what is most important, define what makes for a high-quality objective and provide cues for reflection.

• Align guidance on objective setting with institutional structures. Interventions to encourage high-quality objective setting are more likely to be effective when they complement existing processes, relationships, and ways of working (e.g. approvals processes, budgetary decisions, organisational relationships). This can also mean carefully choosing the messenger and manner of delivery of objectives to align with the context in which they will be received as a way of maximising behaviour change.

• Iterate and test to ensure impact. The impact of changes to the ways in which objectives are set can be hard to anticipate and are likely to have different effects for different parts of government. It is important to measure the quality of objectives and the effect of any interventions made to improve them to make sure they are having the desired behavioural effect.

Why it matters

Public Administrations (PAs) must set objectives to achieve the government’s policy outcomes. Planners must operationalise budget line items into concrete plans with objectives and performance indicators. Project teams must translate long-term vision into medium-term strategy and short-term targets. Civil servants must craft the outputs and outcomes that demonstrate how the government’s intention has been achieved.

However, planning risks becoming an administrative ritual. Objectives may be unambitious or vague. Strategic plans may be rolled over from past years without reflecting new developments. Objectives may not be aligned with public value. Poor planning can be driven by structural factors, such as fragmented responsibilities, but can also be driven by behavioural factors. Civil servants who translate policies into plans may lack the capacity and incentives to do so effectively. This leads to poor objectives setting, creating a gap between administration activity and policy outcomes.

The OECD Behavioural Science Team partnered with the Italian DFP to close this planning gap. The team focussed on improving the quality of organisational objectives set in the Italian public administration’s Piano Integrato di Attività e Organizzazione (PIAO). The PIAO is a comprehensive and consolidated planning document introduced by the Italian Government in 2021 and which entered into force in 2022 to simplify the administrations’ planning process. Administrations must set organisational objectives in their PIAO. The team ran a framed-field experiment to encourage Italian civil servants to select and set high-quality PIAO objectives.

Whom it involves

• Centres of government units responsible for whole-of-government strategy, including guidelines, standards and reporting frameworks for administration plans. Their work motivates, guides and co-ordinates the planning activity of many different public administrations.

• Departmental planning and performance offices responsible for foresight, strategy or long-term planning in their administration. These teams are often responsible the high-level activities, priorities and outcomes the administration seeks to achieve across its areas of responsibility.

What was done

Behavioural techniques can nudge planners to set higher-quality objectives in their administration’s plans. There is high quality evidence that behavioural techniques can help individuals achieve goals – these include techniques described in other chapters, such as pre-mortems, and checklists. However, there is very little published research nudging public administrations, as organisations, to set higher-quality objectives in their strategic plans.

Therefore, the OECD Behavioural Science Team and Italy’s DFP (Department for Public Administration) ran a behavioural pilot to nudge public administrations to set high-quality objectives in their Piano Integrato di Attività e Organizzazione (PIAO). The results of this pilot suggest that behavioural principles can improve how co-ordination units guide and support planning across public administrations, leading to better plans with high-quality objectives. The method and results of this behavioural pilot are described below.

The Piano Integrato di Attività e Organizzazione (PIAO, Integrated Plan of Activities and Organization of Public Administration), entered into force in 2022, is a consolidated planning document all Italian public administrations must publish. In parallel with the implementation of the PIAO, the Italian government developed guidelines, which were subsequently approved on 30 October 2025, to help PAs set high-quality PIAO objectives that promote citizens’ participation, improve transparency and create public value.

Italian civil servants need to overcome behavioural barriers to set high-quality objectives – barriers that can be analysed through the ABCD Framework, which stands for Attention, Beliefs, Choice and Determination (OECD, 2019[1]). Guidance on how to distinguish high-quality objectives from low-quality ones may be more effective if it captures the Attention of civil servants; helps civil servants hold correct Beliefs about what constitutes high-quality objectives as well as how to operationalise them; provides contextual cues that shape civil servant’s Choice to set high-quality objectives and reinforces civil servants’ Determination to set high-quality objectives through motivating factors like deadlines and leadership directives. Each of these behavioural factors shape the setting of PIAO objectives.

Method

The project team conducted a pilot study aimed to encourage civil servants to select and set high-quality PIAO objectives demonstrating public value, as specified in the PIAO guidelines, and help them understand what distinguishes high-quality objectives from low-quality ones. The intervention consisted in a behavioural framed field experiment including three unstratified study arms, one control group and two treatment groups, delivered through a survey on the Qualtrics platform and building on tools already available to DFP for improving objective settings. Treatments were based on the behavioural concepts of:

1. Leadership: One treatment aimed to focus participant’s attention on setting high-quality objectives by signalling that leadership had made it a priority for their PA. Studies have shown that leadership messages can influence civil servants’ behaviour, for example, increasing the likelihood of reporting corruption in the Slovak Republic (OECD, 2024[2]).

2. Feedback: One treatment aimed to shape participants’ beliefs about what constitutes a high-quality objective by helping them understand the characteristics that make it effective. Studies have shown that well-designed feedback can effectively influence behaviour, for example, reducing over-prescription of antibiotics or lowering household energy consumption (Chappell et al., 2021[3]; Kim and Kaemingk, 2021[4]).

3. Simplification: Both treatments aimed to increase the likelihood of civil servants engaging with and acting on the PIAO guidelines by simplifying and clarifying them. Guidelines for setting effective PIAOs, in fact, are often lengthy, and many civil servants may not read them in full.

The experimental sample was drawn from administrators working in Italian PAs. To be eligible, a PA had to be required to upload its PIAO on the Government's PIAO portal and employ at least 50 full-time equivalent staff. This threshold was used because smaller PAs (with less than 50 full-time equivalent staff) are not subject to the same PIAO requirements, and it also reduced the risk of indirect re-identification. Approximately 1 900 PAs met these criteria, collectively representing over 1.2 million full-time equivalent staff. However, the researchers did not expect all eligible PAs to agree to participate.

Recruitment was conducted within PAs on behalf of the research team. The DFP invited designated contact points from PAs with over 50 full-time equivalent staff to distribute an invitation to their colleagues to complete the survey. The invitation included information about the study and an anonymous link to access the survey. The DFP emailed PAs in November 2024, and the survey link remained live from mid-November to mid-December 2024. The research team did not collect participants’ contact details, as recruitment was anonymous and mediated through the internal contact points within each PA. Participants who clicked on the link were first shown an information page outlining the purpose of the survey, its duration, the voluntary and anonymous nature of participation, the identity of the research team, and a contact email for questions. Participants gave their informed consent by entering the name of their PA and proceeding to the next page.

The survey flow was:

1. Participants were presented with a hypothetical scenario describing their PA’s context and the challenges it faces. This served to introduce the context in which they would later be asked to select objectives to include in their PA’s PIAO.

2. After reading the scenario, participants were randomly assigned to one of three study arms: control, leadership, or feedback. The control group proceeded directly to the next section, while the two treatment groups were shown an additional page containing the relevant intervention message, i.e., leadership or feedback. The leadership message was said to be delivered from a PA’s political authority, reiterated the principles of the PIAO guidelines and emphasised the importance of selecting high-quality objectives. The feedback message included high-level guidance on how to set better objectives, building on DFP's draft guidelines. The guidance included an example of a low-quality objective, an explanation of why it did not reflect best practice, and an example of a higher-quality alternative, as well as an example of what a better alternative would look like. Both treatment messages were delivered in Italian. These are show in Table 3.1. Randomisation was handled within the Qualtrics platform and was not visible to participants.

3. All participants then had to rate the quality of eight objectives suggested for their PIAO. Of these objectives, four were high-quality objectives, aligned with the DFPs’ draft guidelines, while the remaining four did not meet these standards (see Table 3.2). Participants were asked to rate the quality of each objective on a scale from 0 to 100. All participants then had to select at least one objective to refine and include in their PIAO, from a list of eight. In both tasks, objectives were presented in a random order to minimise ordering effect. These two tasks aimed to assess participants’ ability to identify and prioritise higher-quality objectives following exposure to the behaviourally informed messages.

4. All participants then answered a set of demographic and background questions. These included their level of satisfaction with the PIAO process, challenges faced in setting objectives, and any feedback they wished to share about PIAO preparation.

5. The survey ended with a debriefing page that thanked participants for their time and provided a contact email for any further questions or concerns.

Table 3.1. Intervention text to improve objective setting

Group	Italian	English
Leadership	Quest’anno sono state pubblicate delle istruzioni per supportare le amministrazioni a sviluppare degli obiettivi meglio definiti e più ambiziosi per il PIAO. Nel trasmettere queste istruzioni agli uffici, il Vertice Politico ha dichiarato: “Obiettivi ben definiti sono quelli in grado di generare Valore Pubblico, cioè benessere economico, sociale e ambientale per i cittadini, le imprese e gli altri stakeholder. Un obiettivo specifico, misurabile e ben definito consente un monitoraggio efficace delle prestazioni e indirizza il personale verso risultati ambiziosi e misurabili. Questo è ciò che ci aspettiamo da tutti i dipendenti pubblici.	This year, the Presidenza del Consiglio dei Ministri has released guidelines to develop better, more ambitious PIAO objectives and reduce the number of unnecessary or superfluous PIAO objectives. To accompany these guidelines, your administration's political authority has said: “Good objectives are about creating public value – the economic, social and environmental well-being of citizens, businesses and other stakeholders. A specific, measurable and well-defined objective allows effective monitoring of performance, and guides staff towards ambitious and measurable results. This is our expectation of all civil servants.
Feedback	Quest’anno sono state pubblicate delle istruzioni per supportare le amministrazioni a sviluppare degli obiettivi meglio definiti e più ambiziosi per il PIAO. Obiettivi ben definiti sono quelli in grado di generare Valore Pubblico, cioè benessere economico, sociale e ambientale per i cittadini, le imprese e gli altri stakeholder. Un obiettivo specifico, misurabile e ben definito consente un monitoraggio efficace delle prestazioni e indirizza il personale verso risultati ambiziosi e misurabili. Ad esempio, "Pubblicare almeno 10 documenti di consultazione" è un obbiettivo mal definito. Questo obiettivo non genera Valore Pubblico, si limita a descrivere un’attività e non è chiaro se 10 documenti rappresentino un miglioramento rispetto al passato. Un obiettivo migliore sarebbe: "Per garantire che le attività dell’Amministrazione siano in linea con le esigenze della comunità, aumentare il numero di consultazioni pubbliche da 10 a 12 all'anno". Le chiediamo di tenere questo a mente quando seleziona gli obiettivi per il PIAO.	This year, the Presidenza del Consiglio dei Ministri has released guidelines to develop better, more ambitious PIAO objectives and reduce the number of unnecessary or superfluous PIAO objectives. Good objectives are about creating public value – the economic, social and environmental well-being of citizens, businesses and other stakeholders. A specific, measurable and well-defined objective allows effective monitoring of performance, and guides staff towards ambitious and measurable results. For example, a bad objective would be: “To release at least 10 consultation papers.” This objective has no clear public value, merely describes activities and it's unclear if 10 papers is an ambitious target. A better objective would be: “To ensure ministry activities align with community needs, we will increase the number of public consultations per year from 10 to 12.” Please keep this in mind when you set PIAO objectives.

Note: In Italian, the Leadership message was 92 words, and the Feedback message was 150 words. Participants were only shown Italian text.

Table 3.2. Fictional objectives that participants chose from

#	Quality	Category	Italian	English
1	High	Public Value	Migliorare l’impatto del programma di certificazione ambientale semplificando la procedura di rilascio e aumentando del 10% gli aderenti, dal 45 al 55%, entro tre anni.	Improve the impact of the environmental certification programme by simplifying the issuing procedure, increasing the number of participants by 10%, from 45% to 55% in 3 years.
2		Service Quality	Aumentare la soddisfazione dei servizi erogati dalla vostra pubblica amministrazione riducendo del 10% i tempi di attesa per i vostri stakeholders.	Increase satisfaction with the services provided by your public administration by reducing waiting times for your stakeholders by 10%.
3		Public Participation	Garantire l’efficacia dell’implementazione del piano sulla digitalizzazione migliorando il processo partecipativo, anche aumentando il numero di incontri con gli stakeholders.	Guarantee the effectiveness of the implementation of the digitalisation plan by improving the participatory process by increasing meetings with stakeholders.
4		Accessible Comms	Migliorare la fruibilità dei documenti pubblicati online, rendendo il linguaggio più accessibile e comprensibile, con un aumento del 10% degli accessi al sito.	Improve the usability of documents published online, making the language more accessible and understandable, with a 10% increase in access to the site.
5	Low	Reduce Headcount	Ridurre l'organico di almeno il 10% del personale equivalente a tempo pieno in tutta l'amministrazione rispetto all'anno fiscale precedente come previsto dalla normativa vigente.	Reduce headcount by at least 10% of full-time equivalent staff across the administration compared to the previous fiscal year as required by current law.
6		Money Saved	Migliorare i saldi di bilancio riducendo la spesa annuale dell’amministrazione del 10% rispetto alla media degli ultimi tre esercizi finanziari.	Improve budget balances by reducing annual government spending by 10% compared to the average of the last three financial years.
7		Comms Published	Pubblicare almeno 10 report o comunicazioni pubbliche che rispettino le linee guida del governo italiano sull'accessibilità e le istruzioni interne circa la identità visiva dell’Amministrazione.	Publish at least 10 reports or public communications that respect the Italian government guidelines on accessibility and internal instructions regarding the visual identity of the Administration.
8		Investments	Incrementare la quota di partecipazione di almeno il 10% in partenariati pubblico-privati nell'area di responsabilità dell'amministrazione rispetto all’esercizio precedente.	Increase participation by at least 10% in public-private partnerships in the administration's area of responsibility compared to the previous year.

Note: Participants were only shown Italian text.

The research team collected a total of 66 450 survey responses. The team conducted pre-specified data quality checks, which led to the exclusion of: Incomplete responses, in which the participant did not answer all mandatory questions. Fast responses, which suggested the participant did not properly read the questions. While most respondents completed the survey in approximately seven minutes, consistent with the estimated completion time, about 2.5% (1 193 participants) completed it in less than 2 minutes and 40 seconds. These responses were removed, as it was unlikely participants had meaningfully engaged with the content. Conversely, nearly 5% of respondents took more than 46 minutes to fill out the survey, possibly because they completed it in multiple sittings. These longer durations were not considered problematic. Non-serious answers, which implied the participant did not provide serious responses. The researchers compared participants’ responses to the questions about their age and the number of years they had worked at their current PA. In 37 cases, participants reported having worked at their PA longer than they’d been alive. These responses were removed, as they suggested the participant may not have taken the survey seriously or had not paid adequate attention to the questions.

After removing low-quality or suspect data, the final sample for analysis comprised 44 193 responses.

Results

Both treatments were associated with a statistically significant change in participants’ objectives selection. The leadership message was associated with a significant increase in the share of respondents selecting the high-quality objective focused on public value. The feedback message was associated with an even larger increase in the share of participants selecting the public value high quality objective. The feedback message was also associated with a decrease in selection of the low-quality objective focused on communications publication. Overall, the feedback message was associated with the greatest positive impact on selecting objectives.

However, there were some unintended consequences. The leadership message was associated with a significant decrease in the share of participants who chose the high-quality objective focused on public participation. The feedback message was also associated with a similar decrease in the share selecting the high-quality public participation objective, and at the same time, with an increase in the share selecting the low-quality objective focused on investments.

The size of the impact was relatively modest. This may have been because participants already chose high-quality objectives. In the control group, the likelihood of choosing one of the four high-quality objectives was 68.3% and the likelihood of choosing a low-quality objective was 25.5%. Furthermore, participants in the control group rated high-quality objectives an average of 70 out of 100 and low-quality objectives an average of 43 out of 100. The interventions may have had less scope to improve upon an already high baseline.

Discussion

The results indicate that while the leadership message was effective, it was less effective than the feedback message. This suggests that civil servants may respond better to specific feedback that includes clear instructions, rather than to general directives to simply set better objectives. PAs therefore benefit from guidance on how to do so, with concrete examples and actionable advice. The feedback message was more effective overall, particularly in increasing the selection of the high-quality objective focused on public value. However, its impact did not extend to other high-quality objectives focused on service quality or accessible communication. This limited effect may have been due to a perceived disconnect between certain objectives (e.g., public participation) and the concept of public value. Also, they may have equated public value primarily with financial investment, overlooking other dimensions. Therefore, providing feedback on how to set public value objectives is necessary but not sufficient. Civil servants in fact benefit from a wide range of well-chosen, illustrative examples.

Behaviourally informed insights

• Behavioural techniques are most effective when aligned with institutional structures. Behavioural approaches can help shape how civil servants set objective by influencing behavioural factors such as attention, beliefs, choice and determination. However, these factors exist alongside structural and institutional conditions, such as performance incentives, approval processes and leadership priorities. Behavioural techniques alone are not a silver bullet, but they can facilitate better planning when they are aligned institutional support for high-quality objective setting.

• Setting high-quality objectives is a multi-stage process. Objective setting in public administrations is a multi-stage process rather than a single decision point. Behavioural techniques may need to be applied at multiple points in the process to have a sustained, positive effect. This pilot intervened at one specific point in the PIAO development process and asked participants to consider a hypothetical scenario when they set draft objectives for later refinement.

References

[3] Chappell, N. et al. (2021), “Using a randomised controlled trial to test the effectiveness of social norms feedback to reduce antibiotic prescribing without increasing inequities”, The New Zealand Medical Journal, Vol. 134/1544.

[4] Kim, J. and M. Kaemingk (2021), “Persisting effects of social norm feedback letters in reducing household electricity usage in Post-Soviet Eastern Europe: A randomized controlled trial”, Journal of Economic Behavior & Organization, Vol. 191, pp. 153-161.

[2] OECD (2024), Improving Corruption Risk Management in the Slovak Republic: Results from a 2023 Experiment in Applying Behavioural Insights to Public Integrity, OECD Publishing, https://doi.org/10.1787/45f8d2e0-en.

[1] OECD (2019), Tools and Ethics for Applied Behavioural Insights: The BASIC Toolkit, OECD Publishing, https://doi.org/10.1787/9ea76a8f-en.