Policy learning from crises: lessons learned from the Italian food stamp programme

Authors:
Simone Busetti University of Teramo, Italy

Search for other papers by Simone Busetti in
Current site
Google Scholar
Close
and
Maria Stella Righettini University of Padova, Italy

Search for other papers by Maria Stella Righettini in
Current site
Google Scholar
Close
Open access
Get eTOC alerts
Rights and permissions Cite this article

Advancing learning is a central tenet for improving public action. Recent calls for agility, robustness, prototyping and other strategies for coping with crises imply continual learning and improvement. This article contributes to challenging this ideal interpretation of the learning process. It provides conceptual and methodological tools to investigate the relationship between policy learning and policy change and sheds light on the diverse dynamics and types of learning that can emerge from crises. At the conceptual level, the article presents a learning matrix that classifies the possible outcomes in the relationship between learning and change. On methods, our research design includes process tracing, binary comparisons, and an innovative real-time approach to the study of learning. The article investigates three municipal case studies from the Italian food stamp programme implemented during the COVID-19 lockdowns. The repetition of the programme over a short period of time offers the opportunity to investigate inter-crisis learning, the process by which lessons from the first wave of implementation contributed to reforms in the second delivery. The coronavirus crisis magnified the acquisition of knowledge and provided radical inter-programme lessons – long-term, non-incremental learning beyond the management of the emergency. Yet, the findings also highlight how this window of opportunity for learning quickly closed and how certain lessons learned may be lost in the process of reform, hard to implement or are unlikely to be extrapolated across contexts.

Abstract

Advancing learning is a central tenet for improving public action. Recent calls for agility, robustness, prototyping and other strategies for coping with crises imply continual learning and improvement. This article contributes to challenging this ideal interpretation of the learning process. It provides conceptual and methodological tools to investigate the relationship between policy learning and policy change and sheds light on the diverse dynamics and types of learning that can emerge from crises. At the conceptual level, the article presents a learning matrix that classifies the possible outcomes in the relationship between learning and change. On methods, our research design includes process tracing, binary comparisons, and an innovative real-time approach to the study of learning. The article investigates three municipal case studies from the Italian food stamp programme implemented during the COVID-19 lockdowns. The repetition of the programme over a short period of time offers the opportunity to investigate inter-crisis learning, the process by which lessons from the first wave of implementation contributed to reforms in the second delivery. The coronavirus crisis magnified the acquisition of knowledge and provided radical inter-programme lessons – long-term, non-incremental learning beyond the management of the emergency. Yet, the findings also highlight how this window of opportunity for learning quickly closed and how certain lessons learned may be lost in the process of reform, hard to implement or are unlikely to be extrapolated across contexts.

Introduction

Learning is fundamental for public administrations dealing with changing contexts and turbulent times, and paradigms for a new public sector place learning centre stage. Agile administrations must engage in continuous self-reflective learning and constantly correct their actions (Mergel et al, 2021), with the concept of experimentalist governance founded on revisions and recursive reviews (Sabel and Zeitlin, 2012). Robust administrations are supposed to learn fast and experiment (Sørensen and Ansell, 2021); their resilience requires policies to be designed considering the complex processes of learning and adaptation (Capano and Woo, 2017). In the same vein, Ansell and colleagues recently proposed prototyping – that is, iterative rounds of testing and revisions – as a promising design strategy to cope with turbulent times (see also Kimbell and Bailey, 2017; Ansell et al, 2020).

These perspectives provide compelling arguments for constantly integrating lessons into revised and more effective public action. They embrace the conventional fascination of rational decision-making inherent in policy learning theories (James and Lodge, 2003), whereby programmes are changed in line with the latest government knowledge. This article contributes to challenging this ideal interpretation. It provides conceptual and methodological tools to study the relationship between policy learning and policy change and sheds light on the diverse dynamics and types of learning that can emerge from managing turbulent problems.

Several concerns can be raised about the relationship between learning and change. On the one hand, learning is fundamentally difficult to measure; lessons are hardly observed in isolation from change (Bennett and Howlett, 1992) and may easily be overstated, missed or just not happen in the timeframe of the research (Radaelli, 2009). On the other, questions also surround the process by which lessons inform changes. Often, updates in government knowledge have no impact on policy formulation (Moyson et al, 2017), public managers’ biases may impede updating policy preferences (Moyson, 2017), and factors related to decision-making – such as focusing events, political dynamics, or veto players (Kingdon, 1984; Birkland, 2006; Tsebelis, 2011; Weible and Sabatier, 2018) – may explain inertia or contribute to changes unrelated to learning.

In crises and turbulent times, the likelihood of policy change following learning is equally mixed. On the one hand, crises confront policymakers with unprecedented challenges, force innovative solutions, and increase the opportunity for radical change (Boin et al, 2017). Crises can be ‘learning triggers’, because they provide experience of new situations and reveal deficiencies of operating procedures (Deverell, 2009). Further, dramatic political and social events may open ‘macro policy windows’ favouring unusual degrees of policy change (Keeler, 1993). On the other hand, however, institutional and political factors in the policy domain (Birkland, 2006; Boin et al, 2008) or cognitive biases of the parties involved may prevent learning from crises and favour inertia (Meyer and Kunreuther, 2017). Rigidity of beliefs, cognitive narrowing, threat minimisation are only some examples of the possible barriers to learning from crises (Smith and Elliott, 2007). In fact, learning organisations need investments, preparation and advanced techniques of information processing and analysis (Lagadec, 1997). The quick pace and short timeframe of a crisis may limit lessons from being processed (Aldrich, 2012) and favour ‘contingent learning’, where changes are first implemented and learning happens only afterwards (Kamkhaji and Radaelli, 2017). The short timeframe and contingent learning regard the intra-crisis period, but can also limit the quality and quantity of the knowledge acquired once a crisis is passed.

This article presents three in-depth case studies about the emergency food stamp programme (FSP) funded by the Italian government and implemented by municipalities in March 2020 during the first COVID-19 lockdown established by the national government. The lockdown entailed restrictions on the mobility of all citizens and the closure of almost all commercial activities. Many workers found themselves without pay, and the emergency started undermining food security. The FSP was initially thought of as a one-time emergency intervention that was not meant to be repeated in the future and whose delivery quickly ended after about a month. In fact, already in May, the lockdown restrictions started to be lifted due to the progressive decrease in COVID-19 cases. However, after about five months with few cases, no restrictions and municipal administrations back to their ordinary activities, the pandemic peaked again in the autumn. Then, at the end of November 2020, the government funded a second delivery of food stamps. These circumstances portray two crises, an inter-crises period and the repetition of the same emergency programme with the same budget in the same context. It offers a unique opportunity to study the process of inter-crises learning – if and how lessons from the first wave of implementation contributed to reforms in the second delivery: Was there learning at the local level? To what extent did lessons from the first delivery improve the implementation of the second FSP? How did the crisis affect this process?

The article proceeds as follows: the first section investigates learning and presents a learning matrix – a conceptual tool portraying how learning may or may not change policies. The second section describes the methodology, and the third section presents the FSP and descriptive data about its implementation in northern Italian municipalities. The fourth to sixth sections investigate the process of learning in the three towns selected as case studies.

Disentangling learning and change

Policy learning is the deliberate adjustment of the goals or techniques of a policy in response to new experiences or information (Hall, 1993). The relationship between policy updates and the use of knowledge has received great attention in the policy literature (Bennett and Howlett, 1992; Sanderson, 2002; Radaelli, 2009; Gerlak and Heikkila, 2011; Heikkila and Gerlak, 2013; Dunlop and Radaelli, 2013; Moyson, 2017; Moyson et al, 2017; Di Giulio and Vecchi, 2019).

Ideally, the conventional process linking learning and change is that represented in Figure 1. A programme’s implementation allows public managers to collect information on the ground and identify unforeseen or unintended effects. Such knowledge needs to be collected (acquisition of knowledge), analysed and translated into policy-relevant lessons (translation of knowledge), and ultimately, contribute to some kind of learning product (Gerlak and Heikkila, 2011; Heikkila and Gerlak, 2013).1

Figure 1:
Figure 1:

The process from learning to change

Citation: Policy & Politics 51, 1; 10.1332/030557321X16678318518550

Source: Adapted from Gerlak and Heikkila (2011), Heikkila and Gerlak (2013)

Throughout this article, we refer to knowledge as raw data and information collected in knowledge acquisition, while lessons indicate knowledge that has been translated into propositions having policy implications. Lessons are the product of information processing and analysis that can derive both from the use of formal methods of production of evidence (Davies et al, 2000), as well as more informal reasoning and discussion on programme implementation. Finally, learning products can take several forms: new strategies, policies, programmes and tools, adjustments to those already in place, or even the decision to maintain or expand them. In other words, learning products may contribute to both ‘social learning’ regarding new policy problems, goals and the scope of action of the administration, and ‘instrumental learning’ regarding the viability of policy tools and implementation design (May, 1992).

Several authors have qualified the impact that learning may actually have, highlighting how there is great variation in the magnitude, depth and quality of changes implemented after learning (May, 1992; Hall, 1993; Levy, 1994; Dolowitz, 2009). Further, as mentioned in the introduction, several barriers may exist when learning from crises. In fact, this ideal learning process is only one of many possibilities.

Table 1 represents a learning matrix – a conceptual tool that portrays these different relationships. The matrix offers a tool for empirical research; it works as an error matrix (Spiegelhalter, 2019: 157) telling when observing or not observing change can correctly predict the presence or absence of learning (Cells 1 and 4) and when these observations can lead to errors (Cells 2 and 3).

Table 1:

Learning matrix

Policy change
Yes No
Policy learning Yes
  • 1. Learning and change

  • i. Within-programme learning

  • ii. Inter-programme learning

  • 2. Learning without change

  • iii. Administrative inertia

  • iv. Confirmatory learning

  • v. Outdated learning

No
  • 3. Change without learning

  • vi. Policy changes independent of learning

  • 4. No change, no learning

  • vii. No policy change or learning

The first cell (learning and change) represents the ideal process of Figure 1: lessons are collected, translated, and then incorporated into programme changes. In this case, observing change correctly predicts the presence of learning (that is, a true positive). Most empirical analyses measure ‘within-programme’ learning (i), when knowledge updates lead changes to the programme being implemented (Birkland, 1997; Brody et al, 2009; Gerlak and Heikkila, 2011; Leach et al, 2014; Moyson, 2017; Di Giulio and Vecchi, 2019). Beyond this, a programme’s implementation may offer lessons that apply more widely to sectors and programmes distant from the source of learning, such as crosscutting knowledge about users or implementation features that can be transferred to other programmes. This kind of inter-programme learning (ii) has received less attention in the literature.

The second cell (learning without change) captures how learning may fail to produce changes. The cell identifies possible false negatives in empirical research; a set of outcomes in which observing the absence of change would incorrectly lead to inferring the absence of learning. The primary reason for this is administrative inertia (iii). Knowledge updates may suggest change, but for several reasons these lessons do not produce changes. Confirmatory (iv) or outdated (v) learning are additional cases where changes are not observed. In the confirmatory case (iv), knowledge updates confirm past choices and suggest maintaining effective features; in the case of outdated learning (v), the administration may have collected knowledge that sets the stage for change, but this has become obsolete by the time the opportunity to reform the programme arrives. This latter situation may be particularly relevant in the case of crises and turbulence, when contexts change rapidly.

The third cell (change without learning), then, regards the possibility of mistakenly assuming that there has been learning when policy changes, that is, a false positive error (vi) (Gerlak and Heikkila, 2011; Heikkila and Gerlak, 2013). As mentioned in the introduction, there are many reasons why policy can change independently of learning, or contrary to the lessons learned, from the action of veto players to that of pressure groups. In the specific case of crisis events, post-crisis politics has been shown to produce a variety of outcomes, in which learning-induced changes are only one possibility (Boin et al, 2008). Often, research documents, interviews and surveys lead the role of learning to be overestimated, when the changes observed have been more contingent on politics and strategy.

Materials and methods

Data

The three case studies draw on multiple sources of information: legislative and administrative documents, newspapers and 12 semi-structured interviews with key informants. The interviews were conducted using video-conferencing, lasted one hour on average, and – when agreed with the interviewee – were recorded, transcribed and analysed by the authors. The social service managers of the three municipalities were the main informants of this work and were interviewed twice, after the first and the second delivery. The interview protocol included in the online Appendix 1 provides details on interviewees and report sample scripts for the semi-structured interviews.

We also present data on the design of the FSP in 43 of the 80 provincial capitals in Italy – those in the following northern regions: Lombardy, Veneto, Friuli Venezia Giulia, Trentino-South Tyrol, Piedmont, Aosta Valley, Emilia-Romagna and Liguria. These data are from official documents issued by the municipalities and focus on three design variables: food aid format, aid amount and eligibility criteria (see Table 3).

Case selection

As mentioned, the crisis and the subsequent implementation of an emergency programme may magnify the potential for acquiring new knowledge. However, several barriers may impede its translation into lessons and learning products. In this regard, although the pandemic provided a tremendous opportunity for knowledge acquisition, the FSP represents a least-likely case for observing learning products (Gerring, 2007). We assume that policymakers and social services had relatively little interest in reforming the FSP. Food aid is not a traditional municipal policy and the programme was designed as a one-off intervention limited to the pandemic emergency. Plus, the programme was a national policy and had a tiny budget compared to most interventions managed by the municipal offices for social services. Finally, we must also note that the time between the first and second waves of delivery was rather short (about eight months), so some ‘learning buds’ may not have had the time to blossom (Radaelli, 2009: 1147). As a result, our hypothesis when entering into this research was that the likelihood of observing learning products in the second delivery was rather low; hence, finding evidence of a process of learning may suggest its prevalence a fortiori in more favourable cases (Gerring, 2007; Beach and Pedersen, 2016).

The three case studies are not comparative but we followed a replication strategy to compile evidence (Yin, 2003). Padua was selected as the first Italian city to deliver food stamps, garnering national attention for its rapid implementation of the programme (RAI, 2020). Vicenza and Verona then implemented the programme in similar contexts and with comparable resources and capacity. The three cities share the same regional welfare system, had never administered food aid programmes before, and are located in the same area (Veneto) – one of the first to be hit by the pandemic, together with Lombardy. The three cities are not politically homogenous; Verona and Vicenza were led by a right-wing government, while Padua had a left-wing governing coalition.

A final comment regards the context for learning. The FSP was implemented and monitored by the municipal social service offices. These are relatively small organisational units made of 2–3 public officials including the social service managers directing the office. Managers directly participated in implementing the FSP and had full discretion in its redesign. In other words, the public officials in the office were the key learners and decision-makers.

Research design

To reliably measure learning, we designed three strands for this research: discrete binary comparison, real-time analysis and process tracing.

First, as shown in Table 2, the Italian municipalities initially designed and administered the FSP in March–April 2020. A second wave of delivery followed between December 2020 and January 2021, with the same budget and legal framework, which provides a rare opportunity to compare two discrete deliveries of the same programme implemented over a short time in the same context. These characteristics, together with the lack of prior experience with food aid, help limit the number of intervening variables in the process of learning leading to change.

Table 2:

Chronology of events and research activities

Date Events and research activities
9 March 2020 First national lockdown. The national government enforces a national lockdown imposing severe restrictions on mobility and closure of commercial activities
29 March 2020 The national government funds the FSP
31 March 2020–9 April 2020 Padua, Vicenza and Verona approve the FSP and start the delivery of food stamps
4 May 2020 Following the development of the pandemic, the national lockdown starts to be relaxed; it will be eliminated in June.
May–September 2020 First phase of the fieldwork: desk analysis, interviews with social service managers, analysis of results about acquisition and translation of knowledge
13 October–3 November 2020 Second national lockdown. Following the new rise of COVID-19 cases, the government progressively implements new restrictions conducive to a new national lockdown
23 November 2020 The national government funds the second delivery of the FSP
9 December 2020–14 January 2021 Padua, Vicenza and Verona approve the second FSP and start the delivery of food stamps
March–July 2021 Second phase of the fieldwork: desk analysis of documents for the second delivery, second round of interviews with social service managers and other interviewees, analysis of results about learning products
Table 3:

Changes implemented between the first and second FSP deliveries

Municipalities Food stamp format Food stamp amount (min/max) Criteria of eligibility
Admission of nonresidents Compatibility with other government subsidies Maximum income Maximum value of savings
Number of municipalities changing design (N=43) 21 (48.8%) 31 (72%) 25 (58.1%) 11 (25.5%) 11 (25.5%) 20 (46.5%)
Padua Yes No Yes No No Yes
Vicenza Yes Yesx Yes No No Yes
Verona Yes Yes Yes Yes Yes Yes

For the second, most empirical analyses investigate learning retrospectively – they observe changes and verify prior learning. This approach is prone to overestimating learning as a change factor, as changes are retrospectively interpreted to have followed lessons and the null hypothesis of no learning is rarely tested (Radaelli, 2009). Similarly, this approach makes it challenging to identify lessons that are ignored since they may easily be downplayed or omitted. Finally, starting with programme changes, retrospective analyses make researchers necessarily biased toward noting within-programme learning.

Our research overcame these issues by using an innovative approach to the study of learning in real time. The research followed the learning process in its chronological unfolding by organising two rounds of fieldwork after each FSP delivery (see Table 2). As aspired by Bennett and Howlett (1992), the first round allowed us to observe learning in isolation, independent of change. It started when COVID-19 cases had dropped and policymakers still thought that the programme had been a one-off emergency intervention not to be repeated in the future. Given these conditions, interviewees could freely present what had worked or not, what they had learned, explore the potential for possible changes, and comment on things that were worth maintaining without feeling compelled to prove that they could make use of knowledge updates. This first step investigated the acquisition of knowledge from the first delivery and its translation into lessons; it was used to draft a list of possible lessons to be used as a benchmark for the second round of the research. The second phase started in March 2021, after municipalities had approved the second FSP. It consisted of a preliminary desk analysis to verify whether the re-design was congruent with the lessons identified in the first phase. We then conducted a second round of interviews and an additional analysis of documents to investigate in more depth why and how lessons were (or were not) applied and whether change or stability actually followed learning. This two-step process allowed us to study learning and change in real time; it provided multiple checks on claims about learning and helped identify omissions, overstatements, changes that did not follow learning, and wider inter-programme lessons.

The final strand of this research regards the use of process tracing to increase the reliability of within-case inferences (Beach and Pedersen, 2019). Though the presentation of the case studies is organised along the cells of the learning matrix, in conducting the fieldwork, we traced the learning process in Figure 1: A) knowledge acquisition, B) translation of knowledge into lessons, C) learning products. The online Appendix 2 provides a summary of these activities and explains when findings were classified as ‘Learning and change’, ‘Learning without change’ or ‘Change without learning’. The case-study sections indicate the codes of the pieces of evidence (for example, A1) reported in the Appendix.

Limitations of this study

The main limit of this study regards external validity. The three cities are very similar organisations, fighting the same crises in similar contexts and cannot be considered representative of the larger group of Northern Italian cities. Our findings can be used in a theory-building way, observing how organisations learn from crises. The three cases allow all the cells in the learning matrix in Table 1 to be populated, offering examples of the different outcomes in the relationship between learning and change.

The Italian food stamp programme

On 29 March 2020, the Italian government allocated about €400 million to municipalities to provide food assistance to their citizens. Funds were allocated based on population sizes and increased by indicators of social deprivation. Municipalities were completely free to design the programmes; they had no guidelines from the government and thus began to experiment with a policy that was completely new to them. On 23 November 2020, the government funded a second delivery with the same budget distribution of the first.

Table 2 shows the 43 provincial capitals in the north that changed the design of the FSP between the first and second waves of delivery and whether Padua, Vicenza and Verona made changes (Yes or No). The analysis considers three design variables: food stamp format, stamp amount and eligibility criteria. These data allow us to describe the implementation of the FSP, but do not reflect whether or not changes followed learning; this will be investigated in the case study sections, and findings are summarised in Table 4 in the concluding section.

Table 4:

Case study findings

PADUA VICENZA VERONA
1) Learning and change
  • No periodic delivery (i)

  • Longer validity (i)

  • Digital FSP (i) and ICT (ii)

  • Housing subsidies (ii)

  • Increased amount (i)

  • Food stamps for ineligible applicants (i)

  • Social supermarket (ii)

  • Centrality of social policy (ii)

  • Wide communication and collaboration with NGOs (i)

  • Decreased amount (i)

  • New eligibility and priority criteria (i–ii)

  • Fund for New Poverties (ii)

2) Learning without change
  • Administrative flexibility (iii)

  • Engagement of volunteers (iii)

  • Targeted actions for vulnerable categories (v)

  • Tailored evaluation (iv)

  • Inclusive criteria (iv)

  • Targexted actions for vulnerable categories (v)

  • Debureaucratisation (iii)

  • Inquiry into vulnerable categories (iii)

  • Targeted actions for vulnerable categories (v)

3) Change without learning
  • New €5000 threshold (vi)

  • Evaluation of applications by area-based social workers (vi)

  • Inclusion of nonresidents (vi)

  • Inclusion of nonresidents (vi)

  • ICT (vi)

  • New €5000 threshold (vi)

  • Inclusion of nonresidents (vi)

  • ICT (vi)

The Italian cities freely experimented with different formats of aid: printed food stamps, prepaid cards, smartphone apps, emailed stamps, or even direct cash transfers. As shown in the table, about half of the municipalities changed the food stamp format between the two deliveries, in particular, by introducing digital stamps. Padua, Vicenza and Verona abandoned the print format for the second wave and instead used a smartphone app, dedicated prepaid card, and the regional health card, respectively.

Concerning the amount of aid, almost two-thirds of the municipalities changed the minimum or maximum value delivered in the second wave. While Padua maintained its original values, Vicenza increased the maximum per family (from €300 to €400) and Verona drastically decreased both the minimum (from €240 to €160) and maximum values (from €800 to €400).

Finally, several municipalities changed the criteria for accessing the programme. The three municipalities in question introduced a threshold for accessing the FSP based on the total value of family savings, and Verona changed other criteria such as the compatibility with other government subsidies and applicants’ income levels. The three municipalities accepted applicants on the second wave without official resident certificates, which were formerly excluded, but this was not a product of learning; it was a legal obligation after several judicial rulings had stated that food stamps were a national and universal programme, the access to which cities could not limit to their official residents only (Actionaid, 2020).

Padua

Padua started to design a municipal food aid programme days before the national government launched the FSP, meaning that it could start to deliver the first food stamps as early as 1 April (Padova Oggi, 2020). In less than three days, the municipality approved the programme, received and evaluated the first applications, printed the food stamps, and organised and started deliveries. The total budget was about €1.1 million, with nearly 4300 families receiving food stamps in the first delivery (Padua municipality, Social Service, email to authors, 27 May 2020).

To apply for the FSP, the residents called an emergency number or emailed, and they were called back by municipal officials to answer interview questions and complete the application (A1, A4). The system was time-consuming, but the municipality was confronted with an unknown condition of need that required in-depth analysis. The office sought to verify an economic need related to the COVID-19 emergency such that applicants could not afford to buy food (Comune di Padova, 2020a). For this reason, the municipality carried out in-depth interviews about the work and family history of all 4271 applicants (Comune di Padova, 2020b) (A1, A4). These interviews provided an unprecedented dataset, analysed and discussed internally by the municipal social service (Interview 1) (B1, B3).

Learning and change

Implementing the programme provided many lessons on how to redesign it (Interview 1) (B4, B5). First, food stamps were initially scheduled to be delivered periodically, but the administration underestimated the number of applications (B4); all the available funds were exhausted with the first delivery. Second, the first food stamps were valid for 15 days, which was considered unsatisfactory (B4): this meant that all the money received had to be spent in a limited time. Finally, the clearest lesson was the need to upgrade the information and communication technology (ICT) of social services (Interviews 1 and 8) (B4). The process of printing, preparing and delivering food stamps, as well as managing the applications one by one, was time-consuming and largely inefficient. Implementing the FSP reinforced the idea that the whole social service needed updated ICT systems (B4, B7).

These lessons were readily applied. During the second delivery, the administration introduced digital food stamps to be downloaded, printed or used through a dedicated smartphone app (Comune di Padova, 2020e) (C1, C2). The digitisation streamlined the activities that had overloaded the administration in April: bureaucratic controls, management of applications, and stamp delivery (Interviews 1, 4, 7 and 8). The digitisation also allowed other changes that were envisaged after the first delivery; food stamps were designed as a one-off intervention with no periodic delivery, and they had a long, one-year validity (Comune di Padova, 2020e) (C1, C2).

The digitisation of the FSP played a fundamental role in a much larger technological change of the municipal social services (Interviews 1, 4, 8) (C1, C2). First, it created a consensus leaning towards digitisation. The first delivery had been so exhausting that everyone in the municipality recognised the importance of upgrading their technology. Second, by starting with a new digital platform for the FSP, rather than a wider project of digital reform, the office avoided a more complex decision-making process, removed possible veto players, and took the first step in a set of incremental changes to expand the FSP platform to other social services (Interview 4).

Finally, the analysis of applicants’ interview responses provided the grounds for making another fundamental inter-programme innovation (B3, B7). Applicants were experiencing major difficulties not only with buying food but also with paying their rent or mortgage. The need to act pre-emptively to avoid a new emergency related to arrears and evictions translated into a new policy goal for the administration (Interviews 1 and 7). Therefore, between October and December, 2020, Padua designed a subsidy for rent and water bills for residents who had suffered losses from the COVID-19 pandemic (Comune di Padova, 2020f) and is now designing an innovative incentive to support landlords and prevent evictions (Interview 4) (C1, C2).

Learning without change

Another lesson learned from the FSP concerned activating both administrative staff and volunteers (Interview 1) (B7). The emergency situation and the need to quickly implement a new programme helped mobilise unexpected resources both within and outside the administration. Internally, there was unexpected flexibility in redirecting employees to implement the FSP; similarly, the mobilisation of citizens was massive, involving about 1500 residents, half of whom had never volunteered before (Comune di Padova, 2020d; Dall’Agnol and Freda, 2020; Interview 7). Both mobilisations were considered potential resources worth maintaining and developing in the future (Interview 1) (B5), but these lessons were not implemented. During the second FSP delivery, most people returned to their jobs and could not volunteer or – in the case of administrative staff – refrained from leaving their routine activities because all municipal services had now been restored (Interviews 4, 7 and 8). In this respect, the new context clearly solidified the prospects for policy change; although the municipality lamented the difficulty in implementing those changes (C4), it took no specific action to maintain the engagement of volunteers or the flexibility of its staff (Interviews 4 and 7).

Another major lesson lost concerned the large grey area of people who – though not in need of assistance in normal times – were vulnerable to even a minimal disruption of their routines (Interviews 1 and 7) (B7). Interestingly, only 10 per cent of the people calling the emergency number during the first delivery had ever contacted social services before the pandemic (Grozny Compasso, 2020) (A3, A4). Most applicants were immigrants (about 75 per cent) who were self-sufficient in normal times, but sent all their savings to their families back home, could not work because of COVID-19 restrictions, and, once left with no earnings, had to contact the social services for first time (Padua Municipality, Social Service, email to authors, 27 May 2020) (A3, A4, B3). The existence of such a grey area highlighted the potential need to monitor these categories of citizens and possibly design a dedicated programme (Interview 1). However, those lessons were quickly outdated because, after the first lockdown, shops had reopened and most such people were no longer in need (further, the national government had provided additional support measures) (Interviews 4 and 7) (C3).

Change without learning

The Social Service Office implemented several changes between the first and second waves of delivery that were not congruent with the lessons learned. During the second delivery, applications were evaluated by the social workers responsible for the area of residence of the applicants. This was certainly an improvement regarding the random allocation used for the first delivery, but it did not result from learning; it was the restoration of the ordinary administrative procedure disrupted by the first wave of the pandemic, when the massive numbers of applicants forced the administration to employ public officials who were not social workers (C5).

The municipality also substituted the more qualitative criterion of ‘not being able to afford food’ for an upper threshold of €5000 in savings. This change was part of a greater standardisation of the programme due to its digitisation and to the will to avoid the case-based evaluations and in-depth interviews of the first delivery (Interview 4). Interestingly, the individual interviews to applicants had been praised as an unprecedented source of information; a lesson missed with the automatisation of applications and the €5000 threshold (C5).

Vicenza

The city approved the programme on 3 April and started to deliver the first food stamps on 6 April. The total budget was almost €600,000, and during the first delivery, involved about 3300 families (Vicenza municipality, Social Service, email to authors, 22 July 2021).

To apply to the FSP, the city set up a digital procedure together with an emergency phone number (Comune di Vicenza, 2020a). With the digital procedure, applicants filled out a web form with information about their job, family, earnings and vulnerability (Comune di Vicenza, 2020b) (A1). For people unable to apply online, the municipality provided a supported procedure through an emergency phone number (A5). Eventually, however, the office contacted all the applicants, even those who had applied digitally (A5). The novelty of the situation was said to require a non-standardised analysis of the conditions of the applicants (Interviews 3 and 9). Near the end of the programme, the digital procedure was even suspended (Città di Vicenza, 2020).

Learning and change

After the first delivery, the social office reported two flaws in the design of the FSP (Interviews 3, 6) (B4). First, food stamps per family were considered too low, especially for large families with young children. Therefore, the municipality redesigned the second delivery by increasing the upper amount from €300 to €400 (Comune di Vicenza, 2020e) (C1, C2).

Second, during the first delivery, about 20 per cent of the applicants worked informally and could not certify their income or prove that they had lost out, which was the only criterion established nationally for applying to the FSP. During the first lockdown, they were redirected from the FSP to the food bags provided by volunteer associations, but this was considered a suboptimal solution (B4). Therefore, in the second delivery, the municipality designed an additional food stamp programme funded through private donations (C1, C2). These food stamps could be delivered to all people considered in need without following the criteria set by the national government; informal workers could now receive more flexible support than only food bags (Interviews 6 and 9).

Interestingly, the main lessons learned from the initial delivery regarded inter-programme learning. The first concerned the possible role of the city in food policy (B7). As mentioned, food aid is normally administered by volunteer associations. However, the emergency limited the participation of volunteers (especially older people) and created several difficulties in delivering food bags. To cope with this special situation, the municipality stepped in, and together with volunteer associations, established a centralised system of delivery to ensure a prompt response and avoid overlaps (A2, A5). This experience was considered worth developing (Interview 3) (B5), and evolved into a partnership aimed at opening a social supermarket providing free food to people in need (Comune di Vicenza, 2021) (C1, C2). Food policy is a completely new sector for the administration and directly derives from learning from the emergency.

A second inter-programme lesson concerned the very role of social policy in the city. The analysis of applications revealed that only 20 per cent of the recipients were already known to social services (Comune di Vicenza, 2020c) (A3, A4). The data showed a large number of people who were routinely self-sufficient but easily vulnerable to fragility. The lesson learned was that fragility may hit virtually anybody, not only the long-time users of social services. Hence, this should be considered more and integrated into all the activities of the municipality (Interviews 3 and 9) (B7). When asked about the implementation of this lesson, the office confirmed an increased role for the social services, as reflected in their increased staff, the design of new services, and the integration of fragility into the overall activity of the administration (Interviews 6 and 9) (C1, C2).

Beyond this, learning from the FSP has been only one element of this transformation. The pandemic generally increased the role of social services, both because the office was one of the few still working during the pandemic (Interview 3) and because the pandemic greatly expanded the budget for social policy, thus enlarging the relative weight of the office (Interviews 6 and 9).

Learning without change

In the case of Vicenza, we did not identify any lessons that remained unimplemented because of inertia, but only some confirmatory and outdated learning.

Concerning the confirmation of past lessons, the eligibility criteria for the FSP identified multiple conditions of fragility and were designed to be as inclusive as possible. The exclusion criteria used in most other social programmes – such as receiving other government benefits – were not considered an obstacle per se; the social workers conducted a tailored evaluation and adjusted the value of food stamps accordingly (Comune di Vicenza, 2020a). Tailoring and inclusiveness were considered worth maintaining (Interviews 3 and 9) (B5) and were confirmed during the second delivery (C1, C2): even people who did not comply with the few eligibility criteria set in the application form were called back for a targeted evaluation (Interviews 6 and 9).

Concerning outdated lessons, during the first delivery, most applicants were suffering job restrictions because of the lockdown, such as workers in restaurants or non-cohabiting caregivers. In our first interview, the administration planned to monitor whether the emergency would be contained in the short run and if dedicated assistance would be needed (B7). During the second delivery, these categories of people were no longer considered at risk and no dedicated programme was designed (C3).

Change without learning

The FSP was completely redesigned as a digital programme, but this lesson was not mentioned during the first round of interviews. The digitisation of the programme certainly meant a major efficiency boost (Interview 9), but when commenting on this digital transformation during the second round of interviews, the social service manager reported how the pandemic had developed a huge market for digital technologies and how the municipalities were constantly contacted by IT companies promoting digital solutions (Interview 6) (C5). In other words, the digital changes to the FSP were not the product of specific learning from the first FSP delivery, but instead, they were partly ‘company-pushed’.

Finally, as in the other municipalities, the second delivery brought some standardisation unrelated to learning. During the first delivery, the municipality set a maximum value of savings to access the programme, with different amounts depending on the number of family members (a maximum of €1000 for each member). This was converted into a single cut-off of €5000, a simplification that streamlined procedures and reduced the number of social workers needed to process the evaluation of applications (C5).

Verona

The city approved the programme on 6 April and started delivering the first food stamps on 9 April. The total budget was almost €1.3 million, with about 3000 families receiving food stamps (Comune di Verona, 2021c).

The procedure required applicants to either call a municipal number or send their request by email (Comune di Verona, 2020a) (A1, A4). The municipal social service established some standard exclusion conditions (such as the simultaneous receipt of other government benefits), as well as a general criterion of eligibility: having experienced a loss of earnings because of the COVID-19 pandemic (Comune di Verona, 2020a). This latter condition was to be verified by social workers, who called all applicants, supported them in completing the application form, and conducted an in-depth inquiry regarding their condition (Comune di Verona, 2020b) (A4).

Learning and change

During the first delivery, the municipality promoted the programme widely, for instance, by putting leaflets in soup kitchens and asking for the support of NGOs to identify potential applicants and help them fill out the application form. Such communication was considered effective in reaching a broad audience (Interview 2) (B4) and was repeated and expanded during the second delivery (C1, C2). In addition to a dedicated phone number, the municipality opened physical counters, and despite setting up a completely digitised procedure, also kept up the possibility of filling out a paper application. Finally, the administration collaborated with NGOs to support specific categories of applicants, such as homeless people and non-resident immigrants (Interviews 2 and 5).

Another change concerned the maximum value of food stamps. In the first delivery, the municipality wanted to give families significant support and decided on an amount per family of up to €800, which had two negative effects (B4). On the one hand, this was too high, considering that food stamps had a short validity; on the other hand, it limited the number of people who could join the FSP. Based on its experience of past emergencies, the municipality estimated that it would receive about 3000 applications, but instead, it received more than 6000 and could eventually approve only 3098 (Comune di Verona, 2020c) (B4). Following this lesson, when designing the second delivery, the office reduced the maximum amount per family from €800 to €400 (Comune di Verona, 2020d) (C1, C2). Interestingly, however, the milder severity of the second lockdown led the number of applications to drop so that the reduction of the maximum amount was probably not needed, at least not as drastically as envisaged after the first delivery (Interview 5) (that is, the lesson was applied, but it was partially outdated).

A major case of inter-programme learning concerned the redesign of the eligibility criteria. Data from the first delivery were analysed and discussed in an internal report (Comune di Verona, 2020c) (B3). Only 27 per cent of the recipients had ever contacted the social services before the FSP (A4). As with the other two municipalities, these numbers revealed a large grey area of poverty and fragility that had been unknown to the social services. In the case of Vicenza, this was interpreted as there being a need to work more on the conditions of eligibility, to avoid excluding vulnerable people from social services (Interview 2) (B7).

A first lesson learned from the analysis of applications was that ‘how much you lost’ was not an effective criterion (B4). There were several cases in which great losses did not mean poverty, whereas small losses entailed serious fragility for others. Instead, ‘how much you were left with’ might have been a better criterion (Interview 2) and was readily introduced in the second delivery: the municipality established income and savings amounts that determined eligibility, even in the case of tiny losses (Comune di Verona, 2020d) (C1, C2). Second, using data from the first FSP applicants, the office ran several simulations, experimenting with different criteria of eligibility (B3), and extrapolated a new set of criteria (Comune di Verona, 2020d) (C1, C2). Interestingly, the new criteria granted eligibility to applicants excluded from the first FSP. These new metrics increased the precision in identifying fragility: they became the new standard for all municipal social programmes and were used in the new delivery of the FSP (Interviews 5 and 12) (C1, C2).

Finally, another inter-programme learning concerned housing subsidies (B7). In the interviews with FSP applicants, difficulties in paying the rent or mortgage emerged as needing immediate action (Interview 2) (A3, A4). During the phone interviews with the applicants, the administration purposefully enquired about the ability to pay the rent/mortgage and utilities (A1, A3). Following the implementation of the FSP, supporting these costs was considered a central policy goal to prevent a future crisis of mass evictions (B6). Accordingly, the municipality established the Fund for New Poverties, a major subsidy programme funded by a local banking foundation that included support for housing (Comune di Verona, 2021b; Fondazione Cariverona, 2021) (C1, C2). This programme had a greater scope than that envisaged following the learning from the first FSP and was favoured also by the persistence of the crisis.

Learning without change

One major lesson learned from implementing the FSP concerned administrative simplifications – the possibility to skip rules, streamline bureaucracy, and speed up procedures (B5). During the first delivery, not only could the FSP applicants self-certify their condition of need without providing any documentation (an expansion of an existing practice) but also no application was subjected to administrative registration. This was a simplification that had never been seen before in the administration and was considered worthy of being replicated in normal times (Interview 2) (B5).

Such de-bureaucratisation was not maintained, however, and all the applications for the second wave of delivery were subjected to administrative registration. Administrative streamlining in ordinary times would have entailed a radical change of all administrative procedures, involving not only the Social Service Office but also the whole of the administration. It was considered too much a radical change, meaning that it was not even attempted by the office – it thus qualifies as a case of inertia (Interviews 5 and 12) (C4).

As in the case of the other two cities, an instance of outdated learning concerned the new categories of fragility to be addressed. According to the analysis of the data by the municipality, some categories of applicants appeared to be especially vulnerable (Comune di Verona, 2020c) (A3, A4). Most recipients were immigrants (54 per cent), and Sri Lankans were highly over-represented regarding their presence in the city (about 17 per cent of all beneficiaries but less than 3 per cent of the total population). The number of single-person households exceeded expectations by far (28 per cent of all beneficiaries), as well as some neighbourhoods, which held the vast majority of recipients (about half of beneficiaries lived in three of the 23 neighbourhoods). These results did not call directly for policy change but rather for further inquiry into the reasons behind the uneven distribution of vulnerability (B7) – an inquiry that nonetheless was not performed by the administration, representing a case of inertia (C4). Finally, the data suggested that workers in the tourist business, especially those employed at nearby Lake Garda, were to be monitored closely in the case of work restrictions continuing until the tourist season (B7). These categories were no longer considered at risk when the second FSP was funded (Interview 5) and thus qualify as cases of outdated learning (C3).

Change without learning

The digital redesign of the FSP could not be related to learning (C5). In designing the first FSP, the administration already considered designing a digital food stamp integrated into other social services of the municipality, but this would have taken a long time (Verona Sera, 2020; Interview 12). The printed food stamps were the quick option in the emergency; digitisation was something that the office would have done in normal times and certainly would have done if the FSP was repeated. When the programme was funded again, the municipality set up a digital platform dedicated to the FSP and integrated with other social services. The digitisation was certainly an improvement, but one that the administration already imagined before the first delivery and so that cannot be considered a learning product.

Discussion and conclusions

Our research provides evidence of how learning unfolds in inter-crises periods and informs subsequent policy change. The case studies present many instances of within- and inter-programme learning, as well as cases of inertia, outdated or confirmatory lessons, and changes independent of learning. Notwithstanding the investigation of three similar cases, the research reveals great diversity in the processes and products of learning. Table 4 summarises the main findings; the remainder of this section discusses the broader results of this research.

The effect of the crisis and the short window of opportunity. The uncertainty and urgency of the pandemic were the major drivers for increasing knowledge. These factors forced the municipalities to use non-routine procedures and case-based evaluations that expanded the quantity and quality of the information collected. Given the broad nature of the crisis, most beneficiaries had never contacted social services, a condition that further increased the novelty of the knowledge acquired. As a result, municipalities could look beyond the long-term users of social services, discover new needs, and ultimately suggest radical changes to social policy. Interestingly, this window of opportunity only characterised the very first wave of the crisis. The second delivery was implemented only a few months later but all municipalities standardised the FSP and partly abandoned the former methods of individual evaluation, reducing the scope for knowledge acquisition. To some extent, this change is inherent in the learning process; experience with the programme leads to standardisation. However, there appears to be a significant trade-off between gains in efficiency and the quality of the knowledge acquired. The potential of the crisis for increasing knowledge can be missed if not met with non-standard tools for knowledge acquisition – tools that are resource-intensive and can be rapidly dismissed.

The centrality of the knowledge translation. Processing knowledge and deriving policy implications is a fundamental stage in the process of learning. The same piece of information may have different interpretations and generate multiple lessons. This was apparent even in cases similar to those examined in this research. For instance, the existence of a grey area of fragile people was interpreted in three different ways: designing a dedicated programme (Padua), integrating fragility into the overall municipal policy (Vicenza), and increasing the precision of eligibility criteria (Verona). The factors affecting these diverse interpretations would need further inquiry and probably have to do with interests, training, previous experience and administrative capacity. However, one point for advancing knowledge translation is worth reminding. In small organisational units like the social service offices, knowledge translation may happen informally through group discussions and individual reasoning on the experience of implementing the programme. However, the most radical lessons seemed to derive from the formal analysis of data and the use of dedicated techniques to produce policy-relevant information. For example, reforming the eligibility criteria in Vicenza was a significant transformation of the service and was possible thanks to simulations and experimentations on FSP data.

The extrapolation of lessons across contexts. One of the main points emerging from our research is that lessons from crises regard special contexts and can be hardly extrapolated outside those contexts. Indeed, instances of outdated learning, such as the identification of new categories of fragility, testify that what administrations learn from managing turbulent problems may rapidly become obsolete even in the next (almost identical) crisis. The erroneous decision to decrease the value of food stamps in Verona is an exemplary case of how lessons can quickly turn out to be outdated, even when implementing the same emergency programme in almost the same context after only a few months. This result qualifies the prospects of applying lessons from turbulent times to more ordinary periods.

The importance of inter-programme learning. Inter-programme learning provides the most radical lessons. Housing support, food security and digitisation are examples of wide-ranging lessons that can give rise to unprecedented non-incremental changes. This suggests that an exclusive research focus on within-programme learning may miss significant increases in government intelligence. Interestingly, inter-programme learning is not only radical but can contribute to long-term changes that – although learnt from crisis – perfectly fit ordinary times. Managers of emergency programmes may have fundamental knowledge beyond the management of the crisis – knowledge that may be important to collect and circulate across the administration.

The path from learning to change. Interestingly, cases of inertia precisely regard such inter-programme lessons, major reforms that the Social Service Office did not even try. Inertia occurred when the changes envisaged after learning were non-incremental and went beyond the sole responsibility of the office managing the FSP. The flexibility of the administrative staff in Padua and the bureaucratic streamlining in Verona are examples of significant disruptions of organisational standards that – although natural during the crisis – became impossible to implement in ordinary times. By contrast, the introduction and reform of housing policy was undoubtedly a radical change, but one under the sole responsibility of the office managing the FSP. As described in the case of the ICT in Padua and the increase in the role of social services in Vicenza, entrepreneurship and political strategy are fundamental to accomplishing non-incremental changes. However, the examples in this research also confirm that the proximity between learners and decision-makers is an effective shortcut in the path from lessons to learning products.

Methods for studying policy learning. Our final comment regards the methodology used in this study. The learning matrix, process tracing, binary comparison and observation of learning in real-time can increase the reliability and accuracy of the empirical investigation of learning. This research design allowed us to overcome ambiguities, omissions, overstatements and faulty memories. For example, an unchanged programme might have been the product of inertia, confirmatory lessons or a lack of learning. Similarly, improvements, such as introducing digital food stamps, were natural candidates for assuming the presence of learning, but they were not necessarily so. In the context of a crisis, programmes may even be consciously implemented in suboptimal ways and improved once times are quieter, without necessarily entailing learning. These are only some examples of the elusive character of policy learning. Although there can still be much uncertainty in recognising whether or not changes and stability follow learning, the article aimed to offer suggestions and good practices for improving research designs for the investigation of learning.

Note

1

Working with collaborative projects, Gerlak and Heikkila also identify the ‘dissemination of knowledge’ as an intermediate step whereby individual lessons are shared and affect practice across the group (Gerlak and Heikkila, 2011; Heikkila and Gerlak, 2013). Yet, this step was not relevant in our case; in the three municipalities, the FSP was managed by the Social Service Office, comprising a small group of bureaucrats who are hierarchically organised and homogenous in terms of training and skills. In other words, collective and individual learning are intertwined in our case.

Conflict of interest

The authors declare that there is no conflict of interest.

References

  • Aldrich, D.P. (2012) Building Resilience: Social Capital in Post-Disaster Recovery, Chicago, IL: The University of Chicago Press, doi: 10.1080/13876988.2013.868660.

    • Search Google Scholar
    • Export Citation
  • Ansell, C., Sørensen, E. and Torfing, J. (2020) The COVID-19 pandemic as a game changer for public administration and leadership? The need for robust governance responses to turbulent problems, 23(7): 94960, doi: 10.1080/14719037.2020.1820272.

    • Search Google Scholar
    • Export Citation
  • Beach, D. and Pedersen, R.B. (2016) Causal Case Study Methods, Ann Arbor, MI: University of Michigan Press.

  • Beach, D. and Pedersen, R.B. (2019) Process-tracing Methods: Foundations and Guidelines, Ann Arbor, MI: University of Michigan Press.

  • Bennett, C.J. and Howlett, M. (1992) The lessons of learning: reconciling theories of policy learning and policy change, Policy Sciences, 25(3): 27594, doi: 10.1007/BF00138786.

    • Search Google Scholar
    • Export Citation
  • Birkland, T.A. (1997) After Disaster: Agenda Setting, Public Policy, and Focusing Events, Washington DC: Georgetown University Press.

  • Birkland, T.A. (2006) Lessons of Disaster: Policy Change After Catastrophic Events, Washington, DC: Georgetown University Press.

  • Boin, A., ’t Hart, P., Stern, E. and Sundelius, B. (2017) The Politics of Crisis Management: Public Leadership Under Pressure, Cambridge: Cambridge University Press.

    • Search Google Scholar
    • Export Citation
  • Boin, A., McConnell, A. and ’t Hart, P. (2008) Governing After Crisis. The Politics of Investigation, Accountability and Learning, Cambridge: Cambridge University Press.

    • Search Google Scholar
    • Export Citation
  • Brody, S.D., Zahran, S., Highfield, W.E., Bernhardt, S.P. and Vedlitz, A. (2009) Policy learning for flood mitigation: a longitudinal assessment of the community rating system in Florida, Risk Analysis, 29(6): 91229, doi: 10.1111/j.1539-6924.2009.01210.x.

    • Search Google Scholar
    • Export Citation
  • Capano, G. and Woo, J.J. (2017) Resilience and robustness in policy design: a critical appraisal, Policy Sciences 50(3): 399426, doi: 10.1007/S11077-016-9273-X.

    • Search Google Scholar
    • Export Citation
  • Davies, H.T.O., Nutley, S.M. and Smith, P.C. (eds) (2000) What Works? Evidence-based Policy and Practice in Public Services, Bristol: The Policy Press.

    • Search Google Scholar
    • Export Citation
  • Deverell, E. (2009) Crises as learning triggers: exploring a conceptual framework of crisis-induced learning, Journal of Contingencies and Crisis Management, 17(3): 17988, doi: 10.1111/J.1468-5973.2009.00578.X.

    • Search Google Scholar
    • Export Citation
  • Di Giulio, M. and Vecchi, G. (2019) Multilevel policy implementation and the where of learning: the case of the information system for school buildings in Italy, Policy Sciences, 52(1): 11935, doi: 10.1007/s11077-018-9326-4.

    • Search Google Scholar
    • Export Citation
  • Dolowitz, D.P. (2009) Learning by observing: surveying the international arena, Policy and Politics, 37(3): 31734, doi: 10.1332/030557309X445636.

    • Search Google Scholar
    • Export Citation
  • Dunlop, C.A. and Radaelli, C.M. (2013) Systematising policy learning: from monolith to dimensions, Political Studies, 61(3): 599619, doi: 10.1111/j.1467-9248.2012.00982.x.

    • Search Google Scholar
    • Export Citation
  • Gerlak, A.K. and Heikkila, T. (2011) Building a theory of learning in collaboratives: evidence from the Everglades restoration program, Journal of Public Administration Research and Theory, 21(4): 61944, doi: 10.1093/jopart/muq089.

    • Search Google Scholar
    • Export Citation
  • Gerring, J. (2007) Case Study Research: Principles and practices. Case Study Research: Principles and Practices, Cambridge: Cambridge University Press, doi: 10.1017/CBO9780511803123.

    • Search Google Scholar
    • Export Citation
  • Hall, P.A. (1993) Policy paradigms, social learning, and the state: the case of economic policymaking in Britain, Comparative Politics, 25(3): 296, doi: 10.2307/422246.

    • Search Google Scholar
    • Export Citation
  • Heikkila, T. and Gerlak, A.K. (2013) Building a conceptual approach to collective learning: lessons for public policy scholars, Policy Studies Journal, 41(3): 484512, doi: 10.1111/psj.12026.

    • Search Google Scholar
    • Export Citation
  • James, O. and Lodge, M. (2003) The limitations of ‘policy transfer’ and ‘lesson drawing’ for public policy research, Political Studies Review, 1(2): 17993, doi: 10.1111/1478–9299.t01-1-00003.

    • Search Google Scholar
    • Export Citation
  • Kamkhaji, J.C. and Radaelli, C.M. (2017) Crisis, learning and policy change in the European Union, Journal of European Public Policy, 24(5): 71434, doi: 10.1080/13501763.2016.1164744.

    • Search Google Scholar
    • Export Citation
  • Keeler, J.T.S. (1993) Opening the window for reform: mandates, crises, and extraordinary policy-making, Comparative Political Studies, 25(4): 43386. doi: 10.1177/0010414093025004002

    • Search Google Scholar
    • Export Citation
  • Kimbell, L. and Bailey, J. (2017) Prototyping and the new spirit of Policy-making, CoDesign, 13(3): 21426, doi: 10.1080/15710882.2017.1355003.

    • Search Google Scholar
    • Export Citation
  • Kingdon, J.W. (1984) Agendas, Alternatives and Public Policies, Boston, MA: Little, Brown and Company.

  • Lagadec, P. (1997) Learning processes for crisis management in complex organizations, Journal of Contingencies and Crisis Management, 5(1): 2431, doi: 10.1111/1468-5973.00034.

    • Search Google Scholar
    • Export Citation
  • Leach, W.D., Weible, C.M., Vince, S.R., Siddiki, S.N. and Calanni, J.C. (2014) Fostering learning through collaboration: knowledge acquisition and belief change in marine aquaculture partnerships, Journal of Public Administration Research and Theory, 24(3): 591622, doi: 10.1093/jopart/mut011.

    • Search Google Scholar
    • Export Citation
  • Levy, J.S. (1994) Learning and foreign policy: sweeping a conceptual minefield, International Organization, 48(2): 279312. doi: 10.1017/S0020818300028198

    • Search Google Scholar
    • Export Citation
  • May, P.J. (1992) Policy learning and failure, Journal of Public Policy, 12(4): 33154. doi: 10.1017/S0143814X00005602

  • Mergel, I., Ganapati, S. and Whitford, A.B. (2021) Agile: a new way of governing, Public Administration Review, 81(1): 16165, doi: 10.1111/PUAR.13202.

    • Search Google Scholar
    • Export Citation
  • Meyer, R. and Kunreuther, H. (2017) The Ostrich Paradox: Why We Underprepare for Disasters, Philadelphia, PA: Wharton Digital Press.

  • Moyson, S. (2017) Cognition and policy change: the consistency of policy learning in the advocacy coalition framework, Policy and Society, 36(2): 32044, doi: 10.1080/14494035.2017.1322259.

    • Search Google Scholar
    • Export Citation
  • Moyson, S., Scholten, P. and Weible, C.M. (2017) Policy learning and policy change: theorizing their relations from different perspectives, Policy and Society, 36(2): 16177, doi: 10.1080/14494035.2017.1331879.

    • Search Google Scholar
    • Export Citation
  • Radaelli, C.M. (2009) Measuring policy learning: regulatory impact assessment in Europe, Journal of European Public Policy, 16(8): 114564, doi: 10.1080/13501760903332647.

    • Search Google Scholar
    • Export Citation
  • Sabel, C.F. and Zeitlin, J. (2012) Experimentalist governance, in D. Levi-Faur (ed) The Oxford Handbook of Governance, Oxford: Oxford University Press, pp 16986, doi: 10.1093/OXFORDHB/9780199560530.013.0012.

    • Search Google Scholar
    • Export Citation
  • Sanderson, I. (2002) Evaluation, policy learning and Evidence-based policy making, Public Administration, 80(1): 122, doi: 10.1111/1467–9299.00292.

    • Search Google Scholar
    • Export Citation
  • Smith, D. and Elliott, D. (2007) Exploring the barriers to learning from crisis organizational learning and crisis, 38(5): 51938, doi: 10.1177/1350507607083205.

    • Search Google Scholar
    • Export Citation
  • Sørensen, E. and Ansell, C. (2021) Towards a concept of political robustness, doi: 10.1177/0032321721999974.

  • Spiegelhalter, D. (2019) The Art of Statistics: Learning from Data, London: Penguin Random House.

  • Tsebelis, G. (2011) Veto Players, Princeton, NJ: Princeton University Press, doi: 10.1515/9781400831456/HTML.

  • Weible, C.M. and Sabatier, P.A. (eds) (2018) Theories of the Policy Process, New York: Routledge.

  • Yin, R.K. (2003) Case Study Research: Design and Methods, London: Sage Publications.

Case-study documents

  • Aldrich, D.P. (2012) Building Resilience: Social Capital in Post-Disaster Recovery, Chicago, IL: The University of Chicago Press, doi: 10.1080/13876988.2013.868660.

    • Search Google Scholar
    • Export Citation
  • Ansell, C., Sørensen, E. and Torfing, J. (2020) The COVID-19 pandemic as a game changer for public administration and leadership? The need for robust governance responses to turbulent problems, 23(7): 94960, doi: 10.1080/14719037.2020.1820272.

    • Search Google Scholar
    • Export Citation
  • Beach, D. and Pedersen, R.B. (2016) Causal Case Study Methods, Ann Arbor, MI: University of Michigan Press.

  • Beach, D. and Pedersen, R.B. (2019) Process-tracing Methods: Foundations and Guidelines, Ann Arbor, MI: University of Michigan Press.

  • Bennett, C.J. and Howlett, M. (1992) The lessons of learning: reconciling theories of policy learning and policy change, Policy Sciences, 25(3): 27594, doi: 10.1007/BF00138786.

    • Search Google Scholar
    • Export Citation
  • Birkland, T.A. (1997) After Disaster: Agenda Setting, Public Policy, and Focusing Events, Washington DC: Georgetown University Press.

  • Birkland, T.A. (2006) Lessons of Disaster: Policy Change After Catastrophic Events, Washington, DC: Georgetown University Press.

  • Boin, A., ’t Hart, P., Stern, E. and Sundelius, B. (2017) The Politics of Crisis Management: Public Leadership Under Pressure, Cambridge: Cambridge University Press.

    • Search Google Scholar
    • Export Citation
  • Boin, A., McConnell, A. and ’t Hart, P. (2008) Governing After Crisis. The Politics of Investigation, Accountability and Learning, Cambridge: Cambridge University Press.

    • Search Google Scholar
    • Export Citation
  • Brody, S.D., Zahran, S., Highfield, W.E., Bernhardt, S.P. and Vedlitz, A. (2009) Policy learning for flood mitigation: a longitudinal assessment of the community rating system in Florida, Risk Analysis, 29(6): 91229, doi: 10.1111/j.1539-6924.2009.01210.x.

    • Search Google Scholar
    • Export Citation
  • Capano, G. and Woo, J.J. (2017) Resilience and robustness in policy design: a critical appraisal, Policy Sciences 50(3): 399426, doi: 10.1007/S11077-016-9273-X.

    • Search Google Scholar
    • Export Citation
  • Davies, H.T.O., Nutley, S.M. and Smith, P.C. (eds) (2000) What Works? Evidence-based Policy and Practice in Public Services, Bristol: The Policy Press.

    • Search Google Scholar
    • Export Citation
  • Deverell, E. (2009) Crises as learning triggers: exploring a conceptual framework of crisis-induced learning, Journal of Contingencies and Crisis Management, 17(3): 17988, doi: 10.1111/J.1468-5973.2009.00578.X.

    • Search Google Scholar
    • Export Citation
  • Di Giulio, M. and Vecchi, G. (2019) Multilevel policy implementation and the where of learning: the case of the information system for school buildings in Italy, Policy Sciences, 52(1): 11935, doi: 10.1007/s11077-018-9326-4.

    • Search Google Scholar
    • Export Citation
  • Dolowitz, D.P. (2009) Learning by observing: surveying the international arena, Policy and Politics, 37(3): 31734, doi: 10.1332/030557309X445636.

    • Search Google Scholar
    • Export Citation
  • Dunlop, C.A. and Radaelli, C.M. (2013) Systematising policy learning: from monolith to dimensions, Political Studies, 61(3): 599619, doi: 10.1111/j.1467-9248.2012.00982.x.

    • Search Google Scholar
    • Export Citation
  • Gerlak, A.K. and Heikkila, T. (2011) Building a theory of learning in collaboratives: evidence from the Everglades restoration program, Journal of Public Administration Research and Theory, 21(4): 61944, doi: 10.1093/jopart/muq089.

    • Search Google Scholar
    • Export Citation
  • Gerring, J. (2007) Case Study Research: Principles and practices. Case Study Research: Principles and Practices, Cambridge: Cambridge University Press, doi: 10.1017/CBO9780511803123.

    • Search Google Scholar
    • Export Citation
  • Hall, P.A. (1993) Policy paradigms, social learning, and the state: the case of economic policymaking in Britain, Comparative Politics, 25(3): 296, doi: 10.2307/422246.

    • Search Google Scholar
    • Export Citation
  • Heikkila, T. and Gerlak, A.K. (2013) Building a conceptual approach to collective learning: lessons for public policy scholars, Policy Studies Journal, 41(3): 484512, doi: 10.1111/psj.12026.

    • Search Google Scholar
    • Export Citation
  • James, O. and Lodge, M. (2003) The limitations of ‘policy transfer’ and ‘lesson drawing’ for public policy research, Political Studies Review, 1(2): 17993, doi: 10.1111/1478–9299.t01-1-00003.

    • Search Google Scholar
    • Export Citation
  • Kamkhaji, J.C. and Radaelli, C.M. (2017) Crisis, learning and policy change in the European Union, Journal of European Public Policy, 24(5): 71434, doi: 10.1080/13501763.2016.1164744.

    • Search Google Scholar
    • Export Citation
  • Keeler, J.T.S. (1993) Opening the window for reform: mandates, crises, and extraordinary policy-making, Comparative Political Studies, 25(4): 43386. doi: 10.1177/0010414093025004002

    • Search Google Scholar
    • Export Citation
  • Kimbell, L. and Bailey, J. (2017) Prototyping and the new spirit of Policy-making, CoDesign, 13(3): 21426, doi: 10.1080/15710882.2017.1355003.

    • Search Google Scholar
    • Export Citation
  • Kingdon, J.W. (1984) Agendas, Alternatives and Public Policies, Boston, MA: Little, Brown and Company.

  • Lagadec, P. (1997) Learning processes for crisis management in complex organizations, Journal of Contingencies and Crisis Management, 5(1): 2431, doi: 10.1111/1468-5973.00034.

    • Search Google Scholar
    • Export Citation
  • Leach, W.D., Weible, C.M., Vince, S.R., Siddiki, S.N. and Calanni, J.C. (2014) Fostering learning through collaboration: knowledge acquisition and belief change in marine aquaculture partnerships, Journal of Public Administration Research and Theory, 24(3): 591622, doi: 10.1093/jopart/mut011.

    • Search Google Scholar
    • Export Citation
  • Levy, J.S. (1994) Learning and foreign policy: sweeping a conceptual minefield, International Organization, 48(2): 279312. doi: 10.1017/S0020818300028198

    • Search Google Scholar
    • Export Citation
  • May, P.J. (1992) Policy learning and failure, Journal of Public Policy, 12(4): 33154. doi: 10.1017/S0143814X00005602

  • Mergel, I., Ganapati, S. and Whitford, A.B. (2021) Agile: a new way of governing, Public Administration Review, 81(1): 16165, doi: 10.1111/PUAR.13202.

    • Search Google Scholar
    • Export Citation
  • Meyer, R. and Kunreuther, H. (2017) The Ostrich Paradox: Why We Underprepare for Disasters, Philadelphia, PA: Wharton Digital Press.

  • Moyson, S. (2017) Cognition and policy change: the consistency of policy learning in the advocacy coalition framework, Policy and Society, 36(2): 32044, doi: 10.1080/14494035.2017.1322259.

    • Search Google Scholar
    • Export Citation
  • Moyson, S., Scholten, P. and Weible, C.M. (2017) Policy learning and policy change: theorizing their relations from different perspectives, Policy and Society, 36(2): 16177, doi: 10.1080/14494035.2017.1331879.

    • Search Google Scholar
    • Export Citation
  • Radaelli, C.M. (2009) Measuring policy learning: regulatory impact assessment in Europe, Journal of European Public Policy, 16(8): 114564, doi: 10.1080/13501760903332647.

    • Search Google Scholar
    • Export Citation
  • Sabel, C.F. and Zeitlin, J. (2012) Experimentalist governance, in D. Levi-Faur (ed) The Oxford Handbook of Governance, Oxford: Oxford University Press, pp 16986, doi: 10.1093/OXFORDHB/9780199560530.013.0012.

    • Search Google Scholar
    • Export Citation
  • Sanderson, I. (2002) Evaluation, policy learning and Evidence-based policy making, Public Administration, 80(1): 122, doi: 10.1111/1467–9299.00292.

    • Search Google Scholar
    • Export Citation
  • Smith, D. and Elliott, D. (2007) Exploring the barriers to learning from crisis organizational learning and crisis, 38(5): 51938, doi: 10.1177/1350507607083205.

    • Search Google Scholar
    • Export Citation
  • Sørensen, E. and Ansell, C. (2021) Towards a concept of political robustness, doi: 10.1177/0032321721999974.

  • Spiegelhalter, D. (2019) The Art of Statistics: Learning from Data, London: Penguin Random House.

  • Tsebelis, G. (2011) Veto Players, Princeton, NJ: Princeton University Press, doi: 10.1515/9781400831456/HTML.

  • Weible, C.M. and Sabatier, P.A. (eds) (2018) Theories of the Policy Process, New York: Routledge.

  • Yin, R.K. (2003) Case Study Research: Design and Methods, London: Sage Publications.

  • ActionAid (2020) La pandemia che affama l’Italia [The pandemic is starving Italy], https://actionaid.imgix.net/uploads/2020/10/AA_Report_Poverta_Alimentare_2020.pdf.

    • Search Google Scholar
    • Export Citation
  • Città di Vicenza (2020) Coronavirus, buoni spesa verso l’esaurimento: le domande si presentano solo al telefono [Coronavirus food stamps: requests to be presented only by phone], Facebook, 30 April, https://www.facebook.com/cittadivicenza/posts/coronavirus-buoni-spesa-verso-lesaurimento-le-domande-si-presentano-solo-al-tele/3749426501765521/.

    • Search Google Scholar
    • Export Citation
  • Comune di Padova (2020a) Misure urgenti di solidarietà alimentare per fronteggiare l’emergenza epidemiologica da Covid-19 [Urgent measures of food solidarity to face the Covid-19 pandemic], 31 March, https://www.padovanet.it/sites/default/files/attachment/Delibera%20G.C.%20n.%2020200143%20DEL%2031032020%20MISURE%20URGENTI%20DI%20SOLIDARIETA%20.pdf.

    • Search Google Scholar
    • Export Citation
  • Comune di Padova (2020b) Bozza di intervista buoni spesa [Interview to FSP applicants], Emailed to the authors, 27 May.

  • Comune di Padova (2020c) Buoni spesa – Elenco supermercati e negozi aderenti [List of supermarkets and shops], 1 April, https://www.padovanet.it/sites/default/files/attachment/supermercati%20buono%20spesa%202304.pdf.

    • Search Google Scholar
    • Export Citation
  • Comune di Padova (2020d) Comunicato stampa: riunione urgente per definire le misure sociali da attivare con contributi del Governo [Press Release: meeting to implement the national government funding], 30 March, https://www.padovanet.it/notizia/20200330/comunicato-stampa-riunione-urgente-definire-le-misure-sociali-da-attivare-con.

    • Search Google Scholar
    • Export Citation
  • Comune di Padova (2020e) Avviso pubblico per l’assegnazione di buoni spesa – seconda erogazione [Public notice to provide food stamps – second delivery], 18 December, https://www.padovanet.it/sites/default/files/attachment/Avviso%20buoni%20Padova%20def%20no%20isplab.pdf.

    • Search Google Scholar
    • Export Citation
  • Comune di Padova (2020f) Avviso pubblico per la raccolta delle domande relative ai contributi del comune di padova emergenza Covid 19: a sostegno dell’affitto e delle utenze dell’acqua [Public call for subsidies supporting rents and water bills], 11 December, https://www.padovanet.it/sites/default/files/attachment/AVVISO%20CONTRIBUTO%20COMUNE%20EMERGENZA%20COVID-PROT_signed.pdf.

    • Search Google Scholar
    • Export Citation
  • Comune di Verona (2020a) Deliberazione Giunta Comunale 118/2020. Risorse da Destinare a misure urgenti di solidarietà alimentare. Provvedimenti [Municipal decision 118/2020. Funding for food solidarity], 6 April.

    • Search Google Scholar
    • Export Citation
  • Comune di Verona (2020b) Coronavirus. Buono spesa, definiti i criteri di assegnazione. Sindaco: Uno strumento in più per nuove categorie di bisognosi [Food Stamps, defined the criteria. The Mayor: A new instrument for new fragile people], 2 April, https://ufficiostampa.comune.verona.it/nqcontent.cfm?a_id=9561&id_com=25848.

    • Search Google Scholar
    • Export Citation
  • Comune di Verona (2020c) ‘Report buoni spesa’ [Food Stamps Report], Emailed to the authors, 31 March.

  • Comune di Verona (2020d) Servizi sociali e integrazione socio sanitaria. Misure urgenti di solidarietà alimentare di cui all’art. 2 del decreto legge 23 novembre 2020 n. 154. Criteri di indirizzo e adempimenti [Urgent measures of food solidarity, art. 2 law 154/2020. Criteria and requirements], 29 December.

    • Search Google Scholar
    • Export Citation
  • Comune di Verona (2021a) ‘Buoni spesa emergenza covid-19 maggio 2021 – proroga termine ultimo per la presentazione delle domande’ [Food stamps May 2021 – Extended deadline], 31 May, https://www.comune.verona.it/media//_ComVR/Cdr/ServiziSociali/Allegati/Buoni%20spesa/PROROGA_6_GIUGNO_2021_PRESENTAZIONE_DOMANDE__DETERMINA_2293_31-05-2021.pdf.

    • Search Google Scholar
    • Export Citation
  • Comune di Verona (2021b) Progetto Una risposta alle Nuove Povertà [Project: A Response to New Poverties], 29 March, https://www.comune.verona.it/nqcontent.cfm?a_id=46989&tt=verona_agid.

    • Search Google Scholar
    • Export Citation
  • Comune di Verona (2021c) Report buoni spesa [Food Stamps Report], Emailed to the authors, 2 August.

  • Comune di Vicenza (2020a) Disposizioni attuative e criteri per la presentazione della domanda di accesso ai ‘buoni spesa’ [Provisions and criteria for receiving food stamps], 3 April.

    • Search Google Scholar
    • Export Citation
  • Comune di Vicenza (2020b) Istanza di accesso ai ‘buoni spesa’ [Application for food stamps], 6 April.

  • Comune di Vicenza (2020c) Coronavirus, modifiche nei criteri per accedere ai buoni spesa [New criteria for receiving food stamps], 15 April, https://www.comune.vicenza.it/albo/notizie.php/251909.

    • Search Google Scholar
    • Export Citation
  • Comune di Vicenza (2020d) Non è più possibile fare richiesta perchè il fondo statale a sostegno dei buoni spesa è esaurito [Closed applications due to budget exhaustion], 5 May, https://www.comune.vicenza.it/servizi/covid19-buonispesa/.

    • Search Google Scholar
    • Export Citation
  • Comune di Vicenza (2020e) Disposizioni attuative e criteri per la presentazione della domanda di accesso ai ‘buoni spesa’ [Provisions and criteria for receiving food stamps], 9 December.

    • Search Google Scholar
    • Export Citation
  • Comune di Vicenza (2021) Emporio solidale vicentino, pubblicato l’avviso per individuare partner del terzo settore [Social supermarket, public call to identify voluntary associations] 26 January, https://www.comune.vicenza.it/albo/notizie.php/270960.

    • Search Google Scholar
    • Export Citation
  • Dall’Agnol, L. and Freda, C. (2020) Padova, Capitale Europea del Volontariato 2020’ [Padua, European Volunteering Capital 2020], Il Mulino, 6 May, https://www.rivistailmulino.it/a/padova-capitale-europea-del-volontariato-2020.

    • Search Google Scholar
    • Export Citation
  • Fondazione Cariverona (2021) Per Verona una risposta alle nuove povertà [In Verona a response to new poverties], 8 March, https://www.fondazionecariverona.org/wp-content/uploads/2021/03/Fondazione-Cariverona-CS_Nuove-poverta-contributo-a-Comune-di-Verona-08-03-2021-DEF.pdf.

    • Search Google Scholar
    • Export Citation
  • Grozny Compasso, I. (2020) Servizi sociali e lockdown: ‘Gli anziani chiamavano per i servizi, i giovani per i buoni spesa’ [Old people called asking for services, young people for food stamps]. Padova Oggi, 31 May, https://www.padovaoggi.it/politica/servizi-sociali-lockdown-anziani-chiamavano-servizi-giovani-buoni-spesa-padova-31-maggio-2020.html.

    • Search Google Scholar
    • Export Citation
  • L’Arena (2020) ‘Verona, buoni spesa a 2.380 famiglie ‘Ma sono in esaurimento’ [2380 families have received food stamps, but they are running low], 9 May, https://www.larena.it/territori/citt%C3%A0/verona-buoni-spesa-br-a-2-380-famiglie-ma-br-sono-in-esaurimento-1.8074422#:~:text=Palazzo%20Barbieri%20ha%20comunicato%20i,milione%20e%20174%20mila%20euro.

    • Search Google Scholar
    • Export Citation
  • PadovaOggi (2020) Buoni spesa, nel pomeriggio inizia la consegna a casa a chi ne ha diritto [Food stamps, in the afternoon the municipality will start the delivery], 1 April, https://www.padovaoggi.it/cronaca/coronavirus-buoni-spesa-comune-padova-01-aprile-2020.html.

    • Search Google Scholar
    • Export Citation
  • RAI (2020) Buoni spesa a Padova per chi è in difficoltà [Food Stamps in Padua for those in need], 2 April, https://www.rai.it/dl/portaleRadio/media/ContentItem-e45dc08a-33bb-4b83-aae5-78e6fe5567bf.html.

    • Search Google Scholar
    • Export Citation
  • Verona Sera (2020) Buoni spesa alle famiglie da 240 a 800 euro [Food stamps to families from €240 to €800], Verona Sera, 2 April, https://www.veronasera.it/cronaca/obbligo-mascherina-coronavirus-virus-buoni-spesa-verona-sboarina-2-aprile-2020.html.

    • Search Google Scholar
    • Export Citation
Simone Busetti University of Teramo, Italy

Search for other papers by Simone Busetti in
Current site
Google Scholar
Close
and
Maria Stella Righettini University of Padova, Italy

Search for other papers by Maria Stella Righettini in
Current site
Google Scholar
Close

Content Metrics

May 2022 onwards Past Year Past 30 Days
Abstract Views 10 0 0
Full Text Views 2295 1182 204
PDF Downloads 1199 533 27

Altmetrics

Dimensions