How policies become best practices: a case study of best practice making in an EU knowledge sharing project

ABSTRACT

Best practices are prevalent in all fields of planning and act to highlight effective and implementable examples, set standards, and generally assist ‘evidence-based' policy-making. In doing so, they frame what futures are desirable and play a role in shaping the planned environment. Despite this power, little is known about how certain policies come to be considered best practices. This article takes a case of best practice making in an EU INTERREG project and illuminates the processes and justifications used to select and formulate best practices. Reviewing project documents and interviewing those involved in selecting possible best practices, demonstrates who decides what should be exemplified, how the decisions are taken, and on what grounds choices are made. The varied and subjective reasonings we find to justify best practices calls into question their perceived neutrality and sturdiness as policy-making instruments. However, selecting best practices, as a process itself, is not without benefits for participants as the reflective element enabled unique forms of learning, opening up wider questions about what function best practices have in making policy.

KEYWORDS:

• Best practice
• policy-making
• policy transfer
• learning
• sustainable transport
• INTERREG

1. Introduction

Land use, transportation, housing, and waste management are just some of the urban issues that call for updated thinking to foster sustainable communities and meet environmental targets. To do so, thorough, effective, and urgent action is required. From a governance standpoint, tackling these issues is complex and entails processes fraught with difficulties, leaving policy makers receptive to information about viable interventions that may assist their work. Policy transfer is as an attractive option to save reinvention of the wheel (Stead, de Jong, and Reinholde 2008) by capitalizing on existing success elsewhere and applying it accordingly.

One mechanism of policy transfer, used to facilitate and expedite the spread of information, is the policy instrument ‘best practice’ (also called ‘good practice’ interchangeably). The underlying principle is replication (Boulanger and Nagorny-Koring 2018). To help address certain challenges, best practices, evidenced as effective ideas delivering results elsewhere, act as blueprints to propel effective policy-making and change. Best practice is popular within the European Union, famously named ‘a massive transfer platform’ (Radaelli 2000, 26), facilitating collaboration between members to encourage the sharing and adoption of best practice (Colomb 2007). The principle being that although cities and regions are unique, the challenges they face are similar, therefore, so are the solutions. Further justification for disseminating and deploying best practices include increasing efficiency and allowing evidence to lead policy decisions. Their ubiquity is unrivalled: ‘no other policy-making instrument has been so widely accepted by policy makers and practitioners’ (Vettoretto 2009, 1068). In this sense, the term has the power to set agendas for what future scenarios are desirable.

Despite widespread usage, best practices are often critiqued. On a fundamental level, what is considered good in one context is not necessarily good in another (Stead 2012, 44) and the de-contextualizing effect of best practice – stripping histories and political climates that enable success – has been problematized (Vettoretto 2009). Furthermore, the deluge of best practices (Stead 2012), the usefulness of them (Wolman and Page 2002), and the way they are circulated (Montero 2017) have each been scrutinized. In terms of how they are used practically, research has moved debate beyond a normative understanding of best practice as a means to replicate policy. Macmillen and Stead (2014) have demonstrated that transport professionals use best practices for a range of ends, particularly as tools to learn from, frame debates, and set possibilities. Elsewhere, it has been observed how best practices can promote a regions green image and city brand (Nagorny-Koring 2019).

These insights are valuable in unpacking best practice policy-making, but what remains absent from the literature is interrogation of how some policies come to don this title. This black-box has been noted (Vettoretto 2009; Stead 2012, 107). By unpacking the processes and decision-making that give birth to best practices this study addresses this gap. With a case of best practice identification and selection in an EU INTERREG project, we shine light on this topic by asking: What processes and justifications do transport professionals deploy to identify and present policies as best practices? By doing so, we illuminate who is (not) involved in the process, what characteristics are considered best practice ‘making’, and detail the effects of searching and reflecting on possible best practices. Understanding such aspects is vitally important, as the best practice moniker sets boundaries of what is and is not exemplified, and hence, is and is not worthy of reproduction.

With little knowledge as to who dictates what becomes best practice and how those decisions are made, this research contributes to more effective and critical use of best practice in policy-making. By scrutinizing the production of best practice, we find multiple reasons are used to justify policies becoming best practices, not simply evidence. This is done through improvised and informal processes which are closely tied and determined by the requirements of the INTEREG project. The process of making best practices, in this case, is shown to provoke an active reflection on policy, acting as a way to learn. These findings sit within a broader debate of best practice and policy transfer, specifically contributing to understanding of the limitations and uses of best practice (Boulanger and Nagorny-Koring 2018).

This paper unfolds as follows: First, we embed the development, spread, and usage of best practice in theory, drawing links from policy studies literature. Second, methods are explained and justified in relation to the CYCLEWALK project – the case study used to explore the development of best practice. Third, findings of the study are presented and discussed, followed by a conclusion which considers wider implications for best practice policy-making.

2. Best practice: usage, conceptualization, and criticism

2.1. Origins and uses of best practice

The term ‘best practice’ was first used by businesses, in the 1960s, looking to improve performance and maintain competitive advantage (Jarrar and Zairi 2000). No one definition of ‘best practice’ exists, but generally the moniker is given to solutions, policies, interventions, actions, or procedures that are deemed successful and may assist other entities grappling with similar challenges (Macmillen and Stead 2014). The term has been applied in many disciplines, including planning and transportation, where it remains standard (Osburn, Caruso, and Wolfensberger 2011, 214).

Examples of best practice are frequently cited in local, national, and international policy documents (see Bulkeley 2006, 131–132; Stead 2012). There are myriad reports and guides charting best practices from local and national governments in the EU, on topics ranging from: pandemic response for fire and rescue services (Fire-In 2020); energy efficiency in data storage centres (EU Science-Hub 2018); to transport topics including freight, road safety, and cycling (TRIMS 2015; SUPREME 2007; ROSE 2014). Between 1997 and 2006, €8.5 billion was spent on EU cross-border projects (Duhr and Nadin 2007, 373–377). More recently, for EU INTERREG alone, €359 million was allocated for projects from 2014 to 2020 (INTERREG 2020). Such inter-regional initiatives ordinarily involve the sharing of best practice to address policy challenges (Colomb 2007), resulting in readily available reports of the sort cited above.

The propagation of best practice can be situated within a wider turn to evidence-based policy-making that occurred in the 1990s. A focus on research and evaluation was set by the 1997 New Labour government in the United Kingdom, who stressed that only proven interventions would become policy (Martin and Sanderson 1999; Sanderson 2002). As part of this effort, the UK government issued guidance in 2001 on how research, both old and new, and consultation with internal and external experts should drive evidenced-based policy (Evans 2009, 249). Embracing evidence opens inquiry to international examples: ‘learning from other countries appears to fit with the notion of ‘evidence-based’ policy-making, a movement in public policy governance that has taken hold in a number of countries’ (Ettelt, Mays, and Nolte 2012, 491). A review of how best practices are deployed concluded that ‘real evidence’ would make them more effective policy instruments (Brannan et al. 2008, 37), showing how evidence-based logic underpins their design. In addition, these trends can be placed amongst heightened financial constraints facing governments across the EU, which increases the appeal of capitalizing on ‘effective solutions established in one area by rolling them out across other areas’ (Marsden 2011, 44).

These factors help explain the proliferation of best practice. Their now ubiquitous status is underpinned by the principle that best practices aid policy-making. This is described as an ‘an accepted wisdom within national policies and programmes, as well as in international arenas and networks’ (Bulkeley 2006, 1030), which stands as testament to the importance afforded to best practice in policy-making processes.

2.2. The transfer and circulation of policy and practice

Policy transfer literature is a helpful starting point to theoretically situate best practice. This work has been described as ‘near-synonymous’ (Macmillen and Stead 2014, 79) with best practice, as it addresses how and why knowledge about policies from one political body are used in another (Dolowitz and Marsh 2000, 5). With their commonly-cited policy transfer framework, Dolowitz and Marsh (1996) provided a means to dissect and analyse the motivations, content, actors, effectiveness, and barriers to policy transfer. Policy goals, content, instruments, programmes, and negative lessons are objects in the process – the what of policy transfer. These concepts can be transferred to differing extents, through direct copying, emulation, hybridization, and inspiration (Dolowitz and Marsh 2000, 9). Best practices package many types of policy and interventions as ideas to be transferred, and are used to inform decisions to various degrees, making these notions of what is transferred and to what extent a valuable conceptual aid.

Partly emanating from critiques of the policy transfer approach, a policy mobilities understanding further contextualizes how best practices are used and circulated. Owing largely to the ‘mobilities turn’ (chiefly Urry 2000; Sheller and Urry 2006), the ‘rational diffusion and best practice replication’ of policy transfer approaches were seen to inadequately frame the ‘politized processes of networking and mutations across shifting social landscapes’ (Peck and Theodore 2010, 17). This perspective argues that the orthodox policy transfer position focuses on actors, but overlooks agency, by assuming that actors are engaged in the rational adoption of an objectively best solution (McCann and Ward 2012, 326–27). Furthermore, mobilities scholars claim that policy transfer is understood and examined too literally, and doing so ignores the complexity of a process where policies are rarely taken ‘off-the-shelf’ in complete form. The process of learning about policy from elsewhere is wedded to how expertize and truth claims are made by transfer agents, with power and politics playing a central role (Temenos and McCann 2012, 11), highlighting that there is ‘nothing natural’ about the policies that are ‘constructed’ as being successful (Ward 2006, 60). As such, this area of research pays close attention to history, political economy, and the scale of movement, adding nuance to the fixity and mobility of particular policies.

2.3. Research and critique

The rather small body of empirical work on best practice draws attention to the ambiguity that exists around their use and reliability. Upon reviewing European sustainable transport best practice examples, Stead (2012) raised concerns over the volume of supposed best practices and the potential information overload burden this could place on policy makers. The quality of the information surrounding best practices has also been questioned, as transport professionals were found to view best practice guides with a degree of scepticism, due to a one-sided fixation on successful accounts of implementation (Marsden 2011, 11). Other research has drawn attention to the fact that, as is often assumed, best practices are not universally applicable because what is ‘best in one place is not necessarily best in another’ (Marsden 2011, 44). This is demonstrated in practice, as contextual differences – cultural set up, language, planning legislation, financial resources and arrangements, etc. – obstructed the uptake of Dutch best practices outside the Netherlands (Pojani and Stead 2015). Re-contextualizing best practices to fit new locales involves adjustments to ‘accommodate the interests of local recipient actors’ (De Jong and Edelenbos 2007, 690), as well as syncing the intervention to another legal, political, and cultural setting, challenging the transferability of the best practice in its original form. Such difficulties mean cherry-picking for best practices that support existing strategy occurs (Ettelt, Mays, and Nolte 2012, 499) as well as only aspects of the example being utilized (Macmillen and Stead 2014, 85).

Research has shown that best practice examples may be less useful for professionals compared with personal contacts, (international) peers, and professional networks (Marsden 2011; Marsden 2011; Wolman and Page 2002, 497). Reasons being that relationships, partnerships, and even informal visits augment the validity and trustworthiness of the information. This evidence suggests that best practices may not play a central role in making policy. Furthermore, the presentation and use of best practice assumes that rational policy makers can select the examples that are most appropriate to them – but this view has been challenged (Winstanley 2012; Wood 2015). For example, in a case study of policy transfer in Guadalajara, Mexico, Montero (2017) finds that best practices from Bogota reach Mexican policy makers only when a coalition of local actors – advocates, politicians, and journalists – mobilized particular examples to influence political decisions (Montero 2017, 346). Seemingly questioning the perceived neutrality of a marketplace of best practices, from which policy makers objectively pick and choose.

Attention is often placed upon how best practices are deployed by practitioners, which contests the normative understanding of best practice, underpinned by the principle that replication valorizes their usage. Boulanger and Nagorny-Koring (2018) note this principle – repeating what is known to work – is held by the EU and its members as a ‘major solution’ to climate-based issues. In contrast, their empirical study of EU projects finds that replication of good policies rarely takes place, partly due to a lack of common understanding of what best practice is and what replication requires (2018, 323). Best practices may not be used as theory intends, but practitioners still utilize them effectively, as Nagorny-Koring (2019) outlines uses beyond just replication: (1) as enablers to provoke stakeholders and politics; (2) as decision-making assists, framing what is doable/desirable; and (3) as currency, to supply organizations with a positive image for place-marketing and public relations. Macmillen and Stead (2014) interviewed transport professionals in the UK and found that although best practices are an accepted part of policy-making, they are viewed in a variety of different ways revealing ‘conceptual ambiguity’ (2014, 85). This demonstrates that best practices are not monolithic and are utilized in different circumstances to different ends. As Macmillen and Stead (2014) recognize, Bulkeley’s (2006) review of usage of best practice resonates with their expanded conceptualization of the policy-making instruments. The ‘accepted wisdom' of best practices is a governmental technology that can affect policy-making discourses (Bulkeley 2006, 1030). Her critical analysis places attention on the inherent authority in the ‘best’ part of best practice.

From existing research, it is possible to derive much about best practices and their role in making policy: why they have become popular devices, the ways in which they are used by policy makers, and how contradictions within their practice challenge their perceived objectivity. However, the processes behind a best practice’s creation, i.e. how they come into being, remains a black box for researchers. Vettoretto (2009) critiques policy transfer literature’s deficiency in explaining why and how best practices are selected, and who decides when a specific example is needed. These claims are echoed by Stead’s (2012, 107) observation that little is known about the production of best practice. By identifying the processes, actors involved, and the types of justifications used to turn a policy into a ‘best practice’, this paper extends the above discussion with outlining how best practices are generated. In doing so, we elucidate a process in which certain policies are identified as examples to learn from, and, as a by-product, others are overlooked. Why and how some ideas become best practices deserves proper attention, and doing so ultimately contributes to the effective learning and adoption of policy from elsewhere.

3. Methods and case study

3.1. Methodology

To investigate our guiding question – what processes and justifications do transport professionals deploy to identify and present policies as best practices? – we use an exploratory case study research design (Yin 2009). This approach allows deep, structured understanding of complex human and social behaviours, in our case to generate insights into best practices (Lewis-Beck, Bryman, and Liao 2004). Two sources of qualitative data were used: semi-structured interviews, conducted in two rounds, and document analysis. Interview participants were partner members of EU INTERREG project CYCLEWALK (2017–2021), which focused on walking and cycling policy and involved the generation of best practices on this topic. Two authors of this study actively participate in the project in an advisory role. It has been demonstrated (see Gray, Laing, and Docherty 2017) that proximity to EU knowledge sharing ventures, through participation, provides a familiarity with project design and structures leading to unique insights. Furthermore, access to participants and project documents enables a thorough methodological approach.

One part of data collection and analysis included project documents submitted by each project partner. These documents outlined and described a selection of policies and interventions partners deemed best practices. Partners selected practices from both their own region and the regions of other partners. For practices from their own region, document data included: (1) practice description; (2) evidence of success; (3) implementation issues; (4) potential for learning/transfer; and (5) additional information. For best practices identified in other regions, less information was necessitated by the project because access to detailed information was likely difficult and, therefore, partners only outlined the practice and explained the potential for learning or transfer. A content analysis was undertaken (Bowen 2009) of these responses which supported the research and development of the interview guide. The documents were formally submitted and gathered in March 2019.

Since the documented responses focused on details of the practices, we required insights into the rationale and process behind them. To do so, semi-structured qualitative interviews were carried out with representatives from each region participating in the project. A semi-structured question guide was developed for participants to express, in their own words, an account of how the best practices came to be (Williams and Best 2014, 247). This allowed for a natural conversation, complementary to the pre-established rapport with some participants due to our participation in the project, shared understanding of each region’s local context, and a mutual grasp of the project’s history. The first round of interviews took place after the practices were submitted (March 2019), and focused on the processes used to complete the activity. A second round of interviews (March 2020) sought not only to validate the previous responses but acquire deeper reflection on processes and justifications behind the decisions. Cumulatively, 15 interviews took place with ten participants who selected best practices (ten project partners and two stakeholders).¹ In two cases, complementary questions were sent via email to validate interpretation and clarify earlier comments. Table 1 outlines interviewees, their professions, and at what stage they were interviewed.² All interviews were carried out using a digital conference calling platform, lasted between 30 and 60 minutes, and were recorded then transcribed prior to coding. Performing a narrative analysis extracted the processes undertaken to produce the potential best practices, followed by an inductive content analysis, using open coding, to identify the reasoning applied to include certain policies and interventions.

Table 1. Research participants.

Name (Pseudonym)	Region	Role in project	Position and organization	Interview Round 1	Interview Round 2
Angela	Burgenland (BUR)	Core partner	Director, regional mobility centre	✓	✓
David	Burgenland (BUR)	Core partner	Transportation planner, consultancy		✓
Laura	Gorizia and Nova Gorizia (EGTC)	Core partner	Communications manager, municipality	✓	✓
Sofia	Gorizia and Nova Gorizia (EGTC)	Stakeholder	Transportation consultant, consultancy		✓
Fabio	Oradea (OLB)	Core partner	Chief engineer, municipality	✓	✓
Beatrice	Olbia (OLB)	Core partner	Chair member, non-profit mobility centre	✓	✓
Paul	Olbia (ORA)	Core partner	Project manager, consultancy	✓	✓
Bruno	Sardinia (SAR)	Core partner	Project manager, municipality	✓	✓
Lukas	Vilnius (VIL)	Core Partner	Head of sustainable mobility, municipality	✓	✓
Tomas	Vilnius (VIL)	Stakeholder	Mobility manager, national ministry		✓

3.2. CYCLEWALK project overview

CYCLEWALK is part of the EU’s INTERREG programme. Projects under this umbrella support regional and local governments across the EU to ‘develop better policy … by creating an environment and opportunities for sharing solutions' (INTERREG 2020). CYCLEWALK (2017–2021) centres on sustainable transportation, with seven partners sharing knowledge on walking and cycling, including infrastructure and policy, in effort to shift mobility behaviours towards these modes of transport. Promoting active modes of transport is a way to reduce societal reliance on cars and alleviate the challenges that come with it, making cycle-friendly policies particularly in demand (Banister 2005; Pucher and Buehler 2017). Within the overarching aims, the project designs to achieve specific outcomes through a series of key activities and deliverables (see Table 2). Six regional governing bodies, either regional authorities, city authorities, or public agencies, responsible for delivering transport policy on a local level, make up the core project partners: Vilnius, Lithuania; Burgenland, Austria; Oradea, Romania; Olbia, Italy; Sardinia, Italy; and Gorizia and Nova Goricia, Italy and Slovenia (border region). The advisory partner is the Urban Cycling Institute, of the University of Amsterdam, supporting the project with technical guidance and scientific knowledge.

Table 2. CYCLEWALK deliverables.

Phase	Key project activities	Description
Phase 1: Initial assessment (2017)	State of Art survey and data collection	A comprehensive survey tool (SOA) to analyse the state of affairs for walking and cycling within each partner region was developed. The categories included (1) Infrastructure, use, design, and guidelines; (2) Financial investment; (3) Maintenance, evaluation, and operations; (4) Communication and education; (5) Policy and vision; (6) Decision-making organizational relations, management, networks; (7) Legal and regulatory framework; and (8) Experimentation and learning.
Phase 2: Active learning (2017–2018)	Study visits	Short trips to each partner region facilitated the study of walking and cycling interventions. Stakeholders from every participant region, especially elected officials, are encouraged to participate. Study visit reports describe what was observed and learned.
Phase 3: Reflection (2019)	Best practice report	To document the collected best practices, a report presents the inter-project best practices.
	Quality criteria tool	A bench-marking framework to assist regions in developing walking and cycling policies is developed from the identified best practices and partner knowledge.
Phase 4: Taking Action (2019–2021)	Local action plans	Each core-partner develops a local plan to outline how knowledge from the project will be implemented, built directly on project experience and other outputs.
On-going	Local stakeholder involvement	At least 50 local stakeholder groups engage with the project through meetings, study visits, and events, and benefit from the activities and collective knowledge.
On-going	Events, workshop, media campaigns, conferences	A series of events and campaigns enable the dissemination of project efforts and knowledge.

Best practice is integral to the project’s design, activities, and execution of deliverables (for an overview see Table 2). CYCLEWALK’s proposal states that ‘regions will reach these objectives (of the project) by selecting and adopting the best practices collected in each other’s region'. No explicit definition of best or good practice is provided on the INTERREG website or within the CYCLEWALK project. One method of collecting knowledge of best practices from other regions was through study visits and project meetings. Project partners and various influential stakeholders joined these study visits. To reflect the collection of ‘best practices’, a core output of the project was a report, systematically codifying practices identified throughout the project.³

To enable the selection and identification of suitable practices for these later objectives, the six regional partners completed an INTERREG-designed template to chart what they considered to be best practices, used for this analysis as described above. The document contained guidance (further clarified by CYCLEWALK partners) on delivering this task, establishing parameters of what could be a best practice, building a contextual definition of the term, and enabling a standardized codification process. First, partners must submit five practices from their own region – ensuring all regions were well-represented. Second, partners were encouraged to submit two practices for each category of walking and cycling based interventions, to again ensure an even distribution of the types of practices selected. These categories were based on a previous deliverable (the state-of-the-art evaluation) and aimed to build a common language to structure approaches to walking and cycling across all deliverables (see Table 2). Third, they were reminded that they could consult experts and stakeholders in their decision-making process, if they wished. Fourth, it was advised to select at least one practice from each of the other regions – again, to ensure even representation.

3.3. Limitations

We outlined the guidance placed on the selection of best practices by the CYCLEWALK project and, owing to this, acknowledge that our findings are bound to the specificities of this particular activity within this particular project. Due to the formal codification of practices in the template, we opted to focus on why and how practices become ‘best’, largely ignoring what policies were not included as best practices. Our analysis provides few specifics of the chosen practices or solutions to retain our focus on the reasons and justifications behind the best practice, without examining merits of the selections made. Furthermore, a vast body of literature documents numerous policies and environmental determinants for increasing active transport.

We also recognize that the best practices were not selected by participants tabula rasa. Although data was collected from participants uniformly, each working from an identical brief, each project partner comes with their own set of experiences, history, culture, language, and professional training and expertize. Additionally, preceding project activities, such as the study visits and project meetings, may have influenced decision-making. Isolating the activity of selecting best practices excludes these foregrounding factors. As a result, our findings do not illuminate cultural or institutional effects and do not include the effects of all project activities.

The relatively small sample of participants hampers the reach of the findings, as seen in other empirical research on best practices in the transport sector (Macmillen and Stead 2014). However, research with practitioners working on transport policy is needed (Marsden and Reardon 2017) and the thick qualitative data presented here opens vital avenues to contemplate best practice making. As has been shown, best practices are found in an array of disciplines and across multiple scales, meaning copious situations produce best practices. The generalizability of our conclusions can be tested alongside other cases of best practice creation, across different fields, in different contexts. We hope our findings advance thinking about the role of best practice, and develop understanding of how certain ideas come to be exemplified as objects that should attract policy makers attention and be learned from.

4. Findings

The findings are presented in two sections. First, the process of creating best practices is dealt with by outlining procedural steps taken by participants and analysing how this process effects the resulting best practices. Second, we chart the reasons which underpin why certain policies and interventions were selected. This allows the reader to understand the chronology of the best practice making process and then, second, the logic which made certain policies best practices; appropriately dissecting both parts of the research question: process and justification.

4.1. Making best practice: process

4.1.1. Instructions and requirements

Each region was provided an INTERREG template outlining the requirements for submitting best practices. Therefore, this document acted as the first point of reference for how to complete the activity and proceed with selection, significantly influencing subsequent actions and decisions:

[the template] was one of the first things that we took into consideration, so we had a look at infrastructure, supporting measures, financing. So, all these issues, and based on that we tried to find out if there is a best practice in the region that will fit to one of these areas. (David, BUR, 19.03.20).

In this sense the document structured and standardized partners’ thinking about best practices, reflecting the structured nature of the form itself. The template also included prompts and suggested word counts for each response, creating a threshold and expectation for descriptions. These elements led some participants to consider what practices ‘we have enough information about’ and ‘would be easier to write’ (Lukas, VIL, 26.03.19). While this rigid outline is essential for uniform completion of the activity, it imposes practical limitations that constrain what is and is not considered a best practice, in effect defining the term. For example, Angela explained that her main take away from her experience in Amsterdam – the social nature of cycling – was not suitable for any category provided by the template (Angela, BUR, 19.03.20). Although the form was not exclusively restrictive. For others, the template was described as ‘very helpful and it also was good for us to go beyond what we thought we could provide as a best practice' (Beatrice, OLB, 24.03.20). In this instance, involving political leaders in EU projects was previously not considered suitable as a best practice but since this action was included in SOA categories, which defined the scope of best practices, it was considered in the final stage.

While this rigid outline is essential for uniform completion of the activity, it imposes practical limitations that constrain what is and is not considered a best practice, affecting the definition of the term. Not only with word counts, but structurally, the activity instructs partners to select a minimum number of examples from particular locales. This restrictive effect was touched upon when discussing the flexibility the form must cater for: ‘you have to consider this form is developed by the INTERREG Europe … It has to fit different themes, it has to fit different types of activities, that’s why it has its own limitations' (Paul, ORA, 20.03.20).

4.1.2. Selecting best practices

Before completing the template, the best practices were chosen and the process to do so differed from region to region. Without specific instructions for how to collect and select best practices, variation may not be surprising but the different processes can be grouped into two main categories. Two partners opted for processes that began with, and were largely driven by, core project partners. Whereas other partners began with external stakeholder input and let the examples they generated lead the process. These stakeholders were already active in the project through organizational affiliation. Project stakeholders are typically made up of related local groups and organizations involved in CYCLEWALK activities. Crucially, both approaches included expert consultation – provided by core project members, close to the overall procedures, who were knowledgeable about partner-wide standards – and stakeholder engagement, but at different stages of the decision-making process.

In project member-led selection processes, discussions among these individuals, during meetings or informal conversations, led to an initial list of possible best practices. This approach was justified on the grounds that those involved had ‘a good overview on all the activities we are running’ (Angela, BUR, 18.03.19) and therefore were well-positioned to oversee the exercise. This type of oversight led to an informal atmosphere at this stage of the activity: ‘There was not really a decision-making process, it was a simple meeting … very easy-going’ (Angela, BUR, 18.03.19). After core project members curated an initial list of possible best practices, stakeholders were then consulted for input. However, this didn’t yield anything ‘very different from what we [herself and the other core project member] thought’ (Beatrice, OLB, 24.03.20). The best practices Bruno selected by himself were raised ‘in meetings with CYCLEWALK stakeholders … and were shared without problems’ (Bruno, SAR, 16.04.20). Consultations with stakeholders were described as a way to ‘finetune’ (David, BUR, 19.03.20) the selected practices.

On the other hand, stakeholder-led selection processes differed in the way ideas were proposed. In meetings, stakeholders were asked to contribute ideas and discuss what may be an eligible policy or intervention for the best practice moniker. In one instance, best practices were discussed across two stakeholder meetings, with the first used as ‘brainstorming – just shovel in whatever practice you consider as being interesting or good' (Paul, ORA, 26.03.19). After collating stakeholder ideas, the core project partners later revised and expanded upon suggestions before completing the template. One partner combined approaches: using a stakeholder meeting to both present the best practice ideas from the core project members and simultaneously collect input from stakeholders (VIL).

Beginning the procedure with stakeholder-led involvement resulted in a formal process, inviting external participants from a variety of organizations, whereas the project member-led approach could be exercised internally and more informally. These variants, which were drawn from the interviews, are outlined below in Table 3.

Table 3. Process of selection.

	Selection process		Nature
	Core partner led	Stakeholder led	Formal decision (meetings with external stakeholders)	Informal decision (discussion or internal casual meeting)
Burgenland	✓			✓
EGTC		✓	✓
Vilnius	✓	✓	✓
Olbia	✓			✓
Oradea		✓	✓
Sardinia	✓			✓

Those that began with stakeholder input reported that opening the search occasionally challenged their initial thoughts on what should be a best practice. Despite having a clear overview of walking and cycling projects in the city, the input of others meant ‘some ideas came up that I didn’t think about' (Lukas, VIL, 26.03.20), which was considered a productive outcome of the exercise. Lukas also stated that meeting stakeholders for this purpose provided space for a positive conversation with members of the cycling community, who are often critical of the administration’s policies, and allowed for a constructive reflection on what is going well in the city. Furthermore, allowing stakeholders to present what they considered to be best practices from their region, in an initial meeting, affected decision-making:

When the stakeholders explained better, we understood that it was to be considered a good practice. At the beginning, I was interested in the, how can I say, the more institutional best practice. (Laura, EGTC, 20.03.20)

Bringing additional voices into the process, outside of the core project members, challenged Laura’s pre-formed assumptions about the types of measures relevant for submission, in this case institutional examples such as inter-municipality governing structures and funding strategies. From a stakeholder perspective on these discussions, Tomas became aware of local interventions which restricted cars and partly-pedestrianized city centre streets, particularly the level of public support for these changes as they are normally strongly opposed. This provided him with a new, local example of best practice to cite in future decision-making (Tomas, VIL, 08.04.20).

4.1.3. Putting practices onto paper

The practices selected then had to be codified in the template provided by INTERREG. Cataloguing their decisions in the detailed manner the template demanded triggered research and reflection. Generally, the prompts on the template were met positively and were described as a ‘good frame’ (Lukas, VIL, 26.03.20) to outline the proposed best practice. One stakeholder, who helped fill in the form, complained that often best practices come with vague descriptions, therefore she appreciated the question that asked for evidence of the practice’s success (Sofia, EGTC, 23.04.20). For one transport consultant, the prompt that required explanation of the ‘potential for learning/transfer’ alerted him to the idea that the practice should be unique and have policy transfer potential, stating ‘this was one of the reasons why things changed from the initial list [of practices]’ (David, BUR, 19.03.20). In short, the template provoked reassessment of an earlier selection.

Completing the form was described, by Bruno, as ‘very simple’ because he was familiar with explaining the effects and difficulties associated with the chosen policies (Bruno, SAR, 16.04.20). Others reported a less straightforward experience due to the template’s rigid requirements. Providing information to properly explain the practice, in some cases, led to internal research, particularly to collect data and evidence of the practice’s success. Laura explained: ‘You have to have all the information, so if you don’t know the practice in detail you can’t complete the boxes. It was the only difficult part’ (Laura, EGTC, 20.03.20). And Paul: ‘It was quite an effort to collect all the data from different departments, we didn’t have data available’ (Paul, ORA, 20.03.20).

Through this research, within their organizations, individuals more closely interrogated the practice. Some identified organizational blind spots: ‘We thought that there should be some kind of monitoring, but there isn’t, maybe this will change. We discussed with the people who are working on this' (David, BUR, 19.03.20). When contacting relevant parties to obtain extra information about a particular practice to complete the form, David learned of extra details of which he was not originally aware. Lukas reported a similar experience when completing the template and collecting information about an online reporting platform used by his municipality to collect citizen feedback on walking and cycling: ‘I was surprised that it was not very expensive … I thought it was way more expensive, and the growth of numbers [of users and reporting] also kind of impressed me’ (Lukas, VIL, 26.03.20). In summarizing the experience of having to complete the template, Lukas described it as ‘a debrief about that project – at last! – which never happened’ (Lukas, VIL, 26.03.20).

4.2. Making best practice: justifications

When explaining the logic and reasoning behind why certain policies were proposed as best practices during interviews, participants cited a variety of justifications to qualify a policy as a best practice. These factors are overlapping and not easily disentangled, with individuals placing emphasis on particular reasons in different circumstances. In the paragraphs below, we explain how these factors are used to justify ‘making’ a best practice. Before addressing these features, it is noteworthy that none of those interviewed reported using an agreed upon method or common definition for what qualifies as a best practice either internally, in their own region, or from regions of other partners. This is exemplified well here: ‘We didn’t adopt some criteria; we didn’t have some scientific criteria. It was only based on our experience’ (Laura, EGTC, 20.03.20).

Evidence of success. Demonstrable evidence of a policy’s success was widely considered appropriate to make it a best practice. While the template reinforced this logic, by explicitly asking for evidence of the policy’s success, ‘success’ or ‘evidence’ were not defined. Many informants corroborated that the inclusion of ‘successful projects, ideas, or successful processes that work in our case’ (Lukas, VIL, 26.03.20) was an important criterion behind selection decisions. Alternatively, a lack of evidence, or poor results, excluded certain policies from the final list: ‘the results are not very good, so although they made investments, it was not selected as a good practice’ (Sofia, EGTC, 23.03.20). Various types of evidence were used, including empirical data: ‘research carried out after the construction of this (cycle) track and some urban cycle paths showed an increase of cycling mobility of 17.3%’ (SAR) and ‘this year the number has increased to 60 pupils from four elementary schools’ (EGTC). Anecdotal and qualitative data also evidenced success: ‘residents were happy with the result’ (VIL) and ‘car traffic was reduced considerably in the city centre’ (ORA). In some cases, secondary data was used to support the policy logic/aims: ‘81% of adolescents do not meet the global minimum recommendations for physical activity’ (BUR). A newly implemented e-bike trial scheme was submitted as a best practice, meaning no data was available on uptake or relative success. During the interview, the inclusion of this practice was justified on the grounds that it addressed a specific local issue and the intervention had been observed as successful internationally (David, BUR, 19.03.20).

A source of pride and spectacle. Policies that people were proud of, wished to exhibit, or that were deemed eye-catching, were also considered appropriate grounds for inclusion. Stakeholders were asked to present ideas on the basis of what ‘you’re proud of, or what you would like to tell the world about’ (Paul, ORA 26.03.19) and others approached the activity by thinking ‘which are the ones we could show … show Europe or the CYCLEWALK partners … those we are really proud of’ (Lukas, VIL, 26.03.20). For one partner a large-scale re-design of a public square, including the pedestrianization and restriction of car access, was described as ‘by far the most spectacular investment, spectacular change in the city in the last several decades’ (Paul, ORA, 20.03.20) and for this reason had to be included in the list.

Uniqueness. The originality of interventions featured as another best practice making quality. For instance, we were told that if a policy is ‘already everywhere – you can be proud that you finally made it – but it’s nothing like a good practice to be shared among the partners and the whole INTERREG programme’ (David, BUR, 19.03.20). The notion that the practice must be unique was considered as a key reason for exemplifying it as best practice, as seen in this explanation of why a stakeholder was opposed to the inclusion of one policy: ‘what they are doing is very good, very important; but there are many other regions with a cycling plan'. (Sofia, EGTC, 23.04.20). Another partner claimed that one policy, where groups of children walk to school along pre-determined routes, was initially ‘not attractive [as a best practice] because it’s a very simple practice in my opinion' (Laura, EGTC, 20.03.20), revealing an assumption that best practices should present their region in an impressive, even sophisticated light.

Applicability. Policies put forward from other regions as best practices were mainly done so with the justification that they could be applied within their own region, i.e. ‘does it fit Vilnius? And can it be useful for us?’ (Lukas, VIL, 26.03.20). However, this approach could also be situated along a broader logic applied to the whole project: ‘from the very beginning, we thought we take exactly what we need for Olbia – also for the identification of good practices’ (Beatrice, OLB, 24.03.20). Regions which are relatable – in terms of size, levels of walking and cycling, geography, etc. – were more likely to have best practices identified from them. To explain why examples were chosen, interviewees revealed: ‘In Amsterdam and Vilnius we saw very good practices but we have some difficulties, many difficulties, to consider them as a practical reference for us’ (Fabio, OLB, 26.03.20). Only two policies from Amsterdam were included; as Angela outlined that Amsterdam was considered a best practice in its entirety and she felt that ‘we cannot just say: Amsterdam’ (Angela, BUR, 18.03.19). While examples from regions with less developed cycling policy were preferred: ‘the Austrian level was more similar to ours. For this reason, it was simpler to apply directly in our territory’ (Laura, EGTC, 20.03.20).

5. Discussion: policy implications and further research

We reflect on our findings across two areas of discussion. First, we consider how further research may advance understanding of best practice policy-making and what can be done to improve their usage practically. Second, we contemplate learning with best practice and what this means for EU projects. The examples that emerge from such projects, and eventually become crystallized as best practice, can direct what our urban futures look like. Therefore, we feel that further exploration of their effects and potential betterment, practically and academically, is a necessary venture.

5.1. Research and practice: bettering best practice

It is recognized that best practice rarely leads to the direct copying of policy (Marsden 2011; Nagorny-Koring 2019), and they are used in various ways, beyond replication (Macmillen and Stead 2014). We build on this understanding by asking: who decided the practice was best, how was it done, and on what grounds were the decisions taken? The web of justifications and processes that make practices best, in the case above, suggests that the entire canon of best practices, cumulatively, have messier and more deeply entangled processes and justifications behind them. To untangle the making of best practice, we outline future research that can build on our findings in breadth and depth, contributing to the fundamental question of how some policies come to set agendas for future actions.

First, broadening this field of research requires extra analysis of the types of scenarios that give birth to best practices. This would not only show what circumstances and routes produce best practices – still a significant blind-spot considering the amount of policies labelled as such – but develop and test our findings. This endeavour might span professions, disciplines, contexts, and cultures, as local planning environments and culturally-specific associations may affect what policies are considered best practices (Ashmore et al. 2018). Specific focus on the procedures that decide best practices, how the inclusion of actors influences outcomes, and what logic underpins application of the best practice label, are all relevant questions. Work of this nature would enable comparison of best practices origins and help build a more complete picture of how these commonly cited policy-making instruments emerge.

Second, research that traces a policy’s transition to becoming a best practice would add needed depth. Our ex-post analysis isolated the completion of an INTERREG-prescribed activity, but, as acknowledged in our limitations, this does not occur in a vacuum. Future analysis would benefit from including exchanges and encounters between actors and sites during projects, taking into account specific actors’ roles, backgrounds, and relationships, which escaped this analysis. Utilizing observational, ethnographic, or multiple-case methods would enable researchers to understand these interactions and incorporate contextual factors. Tracing the evolution of policies may elucidate more varied and complex factors involved, highlighting methods of promotion and presentation, and by whom, revealing additional factors that make some policies become best practice. In-depth work of this nature may require focus on individual policies and projects, studying how they develop, or historically reviewing the trajectories of well-known best practice examples. For this task, policy mobilities research that charts the circulation and history of policy provides cues here (see Bok 2015; Montero 2017; Wood 2015).

Best practice examples, as used by policy makers, would benefit from clear information explaining how it came to be considered so. While the simplicity of the best practice package may account for their ubiquity, asking who, how, and in what scenario would add clarity and transparency, make them more effective, and salvage their potential for genuine benefit. In the absence of a singular vetting authority – subjecting all best practices to identical interrogation and entry criteria – differences will exist, therefore clarity around these areas is needed so their merits can be properly assessed. Although this suggestion would make best practices more sturdy tools, it closely echoes earlier calls that evidence can increase the usefulness of best practices (Brannan et al. 2008). The repetition of this point, while relevant, risks distracting from the beneficial qualities of best practice as is and guards from more fundamental critique.

5.2. Learning with best practice

We found best practices to have varied, and occasionally questionable, reasoning behind their ‘best’ status, but this research also shows that the very activity of identifying, discussing, and researching possible best practices ignited reflection on organizational practices and local policy. This finding contextualizes previous work where practitioner study of best practice examples was considered the most effective method of knowledge sharing within an EU project (O’Dolan and Rye 2012). Our findings suggest that looking inward and searching for best practice in your own locale provoked a particular type of discussion and reflection that unearthed new ideas for those involved. For example, collecting required data for the best practice template led participants to identify organizational short-comings on monitoring. qualitative outcomes of experience and group learning are argued as the true value of knowledge sharing ventures (Duhr and Nadin 2007; Glaser et al. 2020). Such positive accounts of best practice making and usage are likely overlooked due to a tendency to focus on tangible results of EU projects (Colomb 2007, 354).

Our findings suggest best practice creation could be used as a (group) learning exercise, particularly among municipal or regional planners and, sometimes, their key stakeholders. Similar to other strategic learning processes, best practice creation exercises could activate instrumental knowledge about institutional practices and foster capacity-building (Glaser, te Brömmelstroet, and Bertolini 2019); however, this would require more intentional efforts than what was observed in this research. Going forward, EU projects could profit from paying attention to processes, as oppose to current emphasis on policy and outcomes, and explore methods to maximize difficult-to-measure, learning-orientated components of activities. Guidance or options on how to complete activities, who to involve, and what effects this may have could augment learning, with acknowledgment that these processes can play a strategic role, meeting wider sustainability and organizational goals, not only the defined outcome as stated in project proposals. As processes are essential to outcomes, it appears a missed opportunity to ignore how such processes can be optimized. Even so, transport planning’s difficulty with collaboration, communication, and language has been noted (i.e. Bertolini, le Clercq, and Straatemeier 2008) and efforts to work on process would contend with wider disciplinary issues.

The beneficial aspects of the processes we have investigated sit alongside literature that highlights best practices’ potential for learning, rather than their replication (Boulanger and Nagorny-Koring 2018). Here we stress that learning about best practice is not homogenous. It is possible to engage with good policy examples in multiple ways – for instance street infrastructure can be experienced in ways that organizational structures, promotional campaigns, and clean air legislation cannot; suggestions exist on how to incorporate experiential learning into policy transfer activities (Glaser, te Brömmelstroet, and Bertolini 2019; Glaser et al. 2020). If best practice, as a concept, are most aptly used as learning tools it must be considered – by researchers and those responsible for knowledge-sharing projects – what types of engagement enables learning and for what groups. Although this study did not consider how participants learn about the best practices within CYCLEWALK, more understanding of how planners learn through and with best practices would go some way to improving their usefulness, beyond a normative view.

6. Conclusion

This paper explored a gap in recent literature by focusing on the creation of best practice. Our findings indicate that despite their prevalence and common-sense appearance, there is much complexity behind the best practice label. In this case, we found that these agenda-setting instruments are the product of improvised and varying processes of selection and reflection which happen during meetings or informal discussions; are selected by a small groups of experts, or associated bodies who happened to attend meetings; are identified with degrees of evidence, personal experience, and subjective reasoning, without scientific criteria; and are done because of project requirements, which prescribe what can and cannot be exemplified and who can and cannot participate in best practice making.

Documentation and group processes both contributed to how and why practices were selected and considered best. The INTERREG template, with its instructions, definitions, and word-limits, expanded what was thought to be a best practice and excluded certain ideas, revealing how definitions and framing of best practice selection can significantly affect the policies which are (not) chosen. Even when working to uniform requirements, actors used different approaches to select best practices: some let expert knowledge lead the exercise and others consulted stakeholders, to generate ideas from a broader base. When more voices participated in this process, unexpected interventions were proposed, occasionally challenging institutional thinking. This is not to imply including more voices led to different, or ‘better’, outcomes. However, it raises questions about who decides what best practices are, how interests are brought into this conversation, and the resulting effects of this process. Interrogating these processes inadvertently revealed the ways in which the discussion, reflection, and research needed for selecting best practices could benefit organizations and individuals, specifically when searching for best practices in their own region. Collecting this information was not always easy and provoked research and learning about their own organization and its deficiencies. In a sense, this internal micro-audit had worth as a learning process alone. While this allows us to think about the forces behind the exemplification of policies, and what effects they had in our case, these factors remain clouded for most best practices.

It is often assumed that best practices are based on objective evidence but we found this is only partially the case. When evidence is used, it is less than straightforward and takes various forms – qualitative, quantitative, from either primary experience or anecdotal/secondary sources. We do not refute the quality of the evidence presented in our case, nor the relative merits of these types of data generally, but we do point to the divergence: no agreed upon form of evidence was used by those making best practices. Moreover, evidence is not the only criterion to deem a policy best practice. Other notions also contributed to the identification process which show that what is exemplified as best practice is deeply subjective. Ultimately, the grounds which make practices best are not equal and are held to different criteria, yet hold the same title.

Once carrying this title, best practices highlight what policies should be pursued, what planners should strive to implement, and, therefore, what citizens should best live in and through, effectively shaping discourses of future-making. As such, they contribute to the solidification of important moral and political choices with the appearance of a depoliticized, objective instrument (te Brömmelstroet 2020). Our documentation reveals what the authoritative title conceals: the contested and value-laden processes that create best practices. This emphasizes that the transitory process, from policy to best practice, must be thoroughly unpacked to critically reflect on this instrument of policy-making and ask the basic question of: Why are certain futures promoted as desirable and who said so?

To close, we consider the future of best practice policy making, by asking: What do we want best practices to achieve? This can be answered, with an intentionally broad-stroke: best practices should, in some way, advance effective policy-making to address policy issues. Ultimately, only through an understanding of the limits of best practice and what potential they realistically hold – weighing pros against cons – will they most effectively advance planning decisions. If a replication understanding is to be abandoned, as suggested by Boulanger and Nagorny-Koring (2019, 323), then planners must assess their merits primarily as learning tools that hone their individual learning skills and a group’s strategic capacity (Healey 2006). Taking a further step back, the ‘answers are out there’ logic, that underpins best practices, deserves more scrutiny; and this notion may be a more crucial point for critique and reflection going forward. Irrespective of the use best practices have, be it for learning or replication, an overreliance on them risks stymieing radical change and out-of-the-box thinking, and reinforces a status-quo that is detrimental to properly addressing crucial planning issues. Perhaps we should be asking: Are best practices the right way to address urgent challenges?

Acknowledgements

We are grateful to the CYCLEWALK project partners and stakeholders who participated in this research, which was outside the scope of the project activities and deliverables. Additional thanks are to the two anonymous reviewers who provided feedback on an earlier version of this paper. Responsibility for the contents lies solely with the authors and does not necessarily reflect opinions of the European Commission, INTERREG, or CYCLEWALK consortium.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Beatrice and Paul interviewed together to assist with translation to English.

2 Participant names are anonymised throughout the paper.

3 Practices were placed on INTERREG’s open-access ‘Good Practice’ hub: https://www.interregeurope.eu/policylearning/good-practices/.