key: cord-0060280-45kzul6e
authors: Loeffler, Elke; Bovaird, Tony
title: User and Community Co-production of Public Value
date: 2020-07-31
journal: The Palgrave Handbook of Co-Production of Public Services and Outcomes
DOI: 10.1007/978-3-030-53705-0_2
sha: ac547fbbeaeff9da315ad59e17bdc2cf10fa0a0f
doc_id: 60280
cord_uid: 45kzul6e

This chapter considers the many contending definitions of co-production and highlights that they have a common core, containing elements of both citizen action and citizen voice and covering some or all of the Four Co’s—co-commissioning, co-design, co-delivery and co-assessment. From this analysis, the authors develop a definition of co-production and show how it can play a key role in creating public value by improving publicly desired outcomes, making public services more cost-effective, strengthening whole-system resilience and supporting the achievement of public governance principles. A five-step change management model is discussed which helps to operationalise greater and more cost-effective co-production.

and sometimes only involve collaboration in the short-term. So, henceforth, we will use the term 'co-production' to mean 'user and community co-production'.

Again, the rather innocent sentence in the first paragraph above indicates that our interest is in co-production of public services and outcomes (more precisely, publicly desired outcomes). One of the most unfortunate aspects of the New Public Management (NPM) from the 1980s onwards was the tendency to focus on the public sector as simply the commissioner and provider of public services. Yet publicly desired outcomes are often achieved by activities of the public sector which cannot sensibly be called services, e.g. reducing inequality in society. Since most public outcomes require significant contributions of citizens, our interest in co-production extends to how it affects outcomes, as well as simply public services.

We will discuss the meaning of publicly desired outcomes later in this chapter and show how they fit within our public value model, but first we will explore in more detail the definitions of co-production which have been suggested in this Handbook.

We review in Table 2 .1 the wide range of definitions to which reference has been made in the different chapters of this Handbook. (For penetrating analyses of the definitions used in the 'classic' articles on co-production up to the 1980s, see Brudney and England (1983) and Percy (1984) . For an insightful discussion of more recent co-production definitions, see Brandsen and Honingh 2016) . This table mainly focuses on a range of characteristics of co-production highlighted in the literature, including:

• Does the definition of co-production refer to co-production of services or outcomes or both? • Are co-producing citizens just service users or can they also be members of the community? • Is co-production an activity contributing to the work of a public service organisation ('organisation-specific') or is it an activity related more widely to improving a public service or publicly desired outcome? • Does the definition cover all Four Co's (co-commissioning, co-design, co-delivery, co-assessment) or just some of them? • Do the co-producers within public service organisations include the frontline staff, other professionals, managers, politicians and Board members-or just some of these?

The table also highlights other distinctive concepts in each definition which make it unusual. Parks et al. (1981 Parks et al. ( , 1002 'Co-production involves a mixing of the productive efforts of regular and consumer producers. This mixing may occur directly, involving coordinated effort in the same production process, or indirectly through independent, yet related efforts of regular producer and consumer producers' ' The key to efficiency in market arrangements [for co-production] is the capacity of consumers to choose the price and service mix they prefer' (p. 1006)-often not possible with public services Governments may mandate co-productive inputs even where they are not efficient, so co-production is not always voluntary Distinguishes 'regular' and 'consumer' producers Not organisation-specific, rather focused on 'same production process' Mainly focused on co-delivery Service-, not outcome-oriented Focused only on service 'consumers', not community members Brudney and England 1983, 59 ) 'Citizen co-productive activities, or "consumer production", are voluntary efforts of individuals or groups to enhance the quality and/or quantity of services they receive. …

[T]hree types of coproduction are distinguished according to the nature of the benefits achieved: individual, group, and collective' and '… the mix of regular producer and consumer producer inputs that contribute to the production (rather than the destruction) of services' (p. 62) '… co-production requires a "critical mix" of regular producer and consumer (citizen) activities' 'citizens are asked to pitch in and help insure the quality of life in their city' (p. 59) '… coproduction involves voluntary, cooperative action in service delivery'

Distinguishes individual from group and collective co-production Focused on community members as well as service 'consumers' Outcome-as well as service-oriented Focuses particularly on co-delivery but also accepts co-commissioning ('transmission of preferences') and co-assessment Not clear what how 'critical' is assessed (continued) Alford (2009, 23) 'Co-production is any active behaviour by anyone outside the government agency which:

• is conjoint with agency production or independent of it but prompted by some action of the agency PSO therefore actively drives the value co-creation process by inviting service users to engage in service production through co-production' Also define kindred concepts of 'co-experience', 'co-construction'

and 'co-design' which overlap with many aspects of co-production Contributions by citizens must be voluntary Differentiates two intrinsic (co-experience and co-construction) and two extrinsic (co-production and co-design) processes of participation, giving attention to value created in service use, not simply in service production Sees public service staff controlling co-production-focuses on 'inside-out' co-production It has long been understood that, in many services, the acts of production and consumption are essentially inseparable, requiring the simultaneous contributions of both producers and consumers, so that some level of coproduction is inevitable. However, that is not the issue which the authors in this book are addressing-they are concerned rather with the question of what is the optimal mix of the contributions of each of the stakeholders-producers, consumers and 'co-producers'-since there is almost always a range of choices available for this mix. This was the original question posed by the scholars in the Bloomington School and each of the definitions in Table 2 .1 can be seen as ways of limiting the scope of inquiry to make this complex problem more tractable.

It will be seen from Table 2 .1 that a number of authors use the phrase 'cocreation' for some aspects of co-production, as we have defined it. A valuable commentary on the literature by Voorberg et al. (2015 Voorberg et al. ( , 1347 summarises this debate: '… we concluded that in the literature the concepts of co-creation and co-production were often seen as interchangeable. There is empirically no striking difference between both concepts, and within bodies of knowledge different meanings are given to both concepts'. We feel this is a fair assessment of the current situation, so we do not give a separate definition of 'co-creation', although we recognise that the phrase 'co-creation of value' is often used to encapsulate the overall approach to using the Four Co's to improve outcomes. Table 2 .1 also makes clear that the term 'citizens' occurs in many definitions, without any attempt to define clearly the basis of this 'citizenship'-e.g. does it apply only to residents with formal citizen status, or to all residents, or …? Since all co-production definitions are consistent with different citizenship concepts, we do not consider this issue further in the Handbook.

One distinction from the literature which we have not coded in Table 2 .1 is between 'instrumental' and 'transformative' co-production. As discussed in Chapter 1, the issue of whether co-production can bring about fully transformative change in society and citizen experiences or whether it is generally only instrumental in its effects is a theme in several chapters of this Handbook. However, as we have noted elsewhere (Loeffler and Bovaird 2016, 9) , the organisational literature on 'transformation' is highly contested: there are few rigorous definitions of what constitutes 'transformation', no clarity on the relevant time scale within which change might qualify as 'transformational', no agreement on the criteria for when 'transformation' can be said to have occurred, and a huge variety of drivers or preconditions suggested as necessary for 'transformational change'. Moreover, while many authors suggest co-production offers the potential for transformative change, it is noticeable that not many examples of 'transformation' of outcomes are presented in these chapters (with the exceptions of Chapters 16, 21, 28, 29 and 30)-and even where it is suggested to have occurred, it is generally not demonstrated that user and community co-production caused it. Indeed, where radical improvements in service outcomes and costs have indeed occurred, as for example in the Surrey CC Services for Young People (Bovaird and Loeffler 2014) , many contributory factors in addition to co-production can usually be identified. Finally, the operational value of this distinction is unclear-it is generally not possible to know in advance that the inputs from the co-producers will be sufficient to achieve transformation-at least in complicated, complex and chaotic knowledge domains-so this distinction usually cannot be made a priori. Taking all these points into consideration, we regard this distinction between 'instrumental' and 'transformative' co-production to be generally unhelpful. Table 2 .1 also includes three different definitions associated with the authors of this chapter. This could be seen, on the one hand, as overenthusiasm for writing definitions or, on the other hand, as an inability to make up our minds. However, our journey from the 2007 definition to the 2020 definition may be of some interest. The original (Bovaird 2007) definition has been widely cited in the academic literature but we quickly became convinced that it was incomplete-in particular, it excluded short-term coproduction activities, even when they might be of real significance, and it focused on public services, rather than outcomes-part of the UK tendency to conceptualise everything in the public domain as a 'service'. Moreover, in common with many of these definitions, it conveyed little to policymakers or practitioners. Consequently, from 2009 onwards we developed a different approach to defining co-production-this is represented in Table 2 .1 by the Bovaird and Loeffler (2013a) definition. This avoided the disadvantages of the Bovaird (2007) definition, highlighting that co-production was about the effect of both citizens and professionals in co-production on outcomes (or costs), whether in short or long term. This definition also has the advantage of focusing on a joint optimisation process-co-production is about the potential for both sides to benefit more thoroughly from each other's assets, resources, capabilities and contributions.

We have used this definition in our research, publications and training ever since. However, we have slowly become uneasy about this definition focusing purely on 'professionals' as the contributors from the public service organisation side of co-production activities, when clearly others might well be involved, including elected politicians in government organisations, Board members of non-profit organisations, managers and, perhaps most of all, frontline staff without professional qualifications. Consequently, Loeffler (2021) has made a slight change to our previous definition and this is the definition which lies behind the rest of this chapter and the final chapter of the Handbook:

Public service organizations and citizens making better use of each other's assets, resources and contributions to achieve better outcomes or improved efficiency.

However, the key message from this section is not our own preferred definition but rather that definitions depend on the purpose to which they are put-we agree with Brudney and England (1983, 61 ) when they suggest: 'As with any definition, the issue is not whether the interpretations of co-production found in the literature are valid, but whether they are useful'.

Consequently, we believe it is not sensible to seek to impose a single definition on the field. Rather, public service organisations should consider carefully which definition suits them (not simply taking a common definition 'off the shelf'). From that point onwards, they should state clearly what they mean by 'co-production', make sure that this definition is known to those dealing with them, and stick to that definition-but understand that others may be using a different definition. At some stage, of course, they may find it useful to reconsider and change the definition they use-then the cycle starts again. In that way, we believe, it is possible to use language to communicate, not to dominate.

Co-production is not undertaken for its own sake. As discussed in the previous section, it can be seen to be either a pathway to improving public services or publicly valued outcomes. We therefore need a conceptual framework within which to locate co-production. In Fig. 2 .1 we present the Governance International Public Value Model, which demonstrates a range of pathways by which co-production can add public value.

In this model, we define public value as the balance between the achievement of priority public quality of life outcomes and priority public governance principles (Loeffler and Bovaird 2019) . This is a different definition of public value from that used by Moore (1995) , but our Public Value Model is consistent with Moore's 'strategic triangle', while giving more detail and being more operational. In Moore's approach, public value is created through the production of services valued by stakeholders (at low cost), within a system which provides legitimacy and political support, and also ensures the operational and administrative capacity of the public services system (Moore 1995, 71) . Our Public Value Model in Fig. 2 .1 contains elements which address each of these concepts-public services valued by citizens are commissioned by the public sector, legitimacy and political support are specific examples of wider public governance principles which are imposed on the whole system, and the capacity of public service organisations determines how well their services provide outcomes to different stakeholders.

However, our Public Value Model goes further to highlight explicitly that the expressed demands of different stakeholders must be politically mediated to arrive at publicly defined 'needs'. Some of these needs are met by the public services which are commissioned by the public sector but which can be provided by public, private or third sector provider organisations, or from partnerships between them. In Fig. 2 .1 the horizontal arrows from providers show how these commissioned services produce quality-of-life outcomes for individual service users (user value), communities (social value) and for businesses (economic value). This is the traditional pathway to outcomes assumed by most public managers and policymakers. However, the Model makes clear that prioritised needs can be met not only by providers of commissioned public services but also by citizen behaviour change or their co-production with public service organisations. For example, if people are encouraged to adopt a healthier lifestyle, they are likely to stay healthy longer and, as a result of this behaviour change, fewer health and care services will be required. The same result may eventuate if people who are ex-substance abusers co-produce by giving peer group support (alongside inputs by public service organisations such as training of peer supporters) to people who are attempting to rehabilitate. In these ways, citizens who change their behaviour or take part in co-production activities are likely to increase their self-confidence and self-esteem-and in the case of co-production, also to improve their own level of social contact and feel less isolated. All of these outcomes are valuable in themselves, although not directly part of the commissioned services which they may also be helping to co-produce. Such pathways are shown in Fig. 2 .1 by the vertical arrows to the quality-of-life outcomes desired by service users, communities and the business sector.

Moreover, the Model highlights how behaviour change and user and community co-production are often intertwined in a close and complex interrelationship, in which both can be the cause or the effect of the other. Where behaviour change is encouraged and enabled by public service organisations, it is itself a form of co-production-this happens, for example, where substance abusers respond to treatment and reduce their level of dependence. Where behaviour change is self-driven by service users or community members, it can have the positive effect of increasing the ability of those citizens to contribute to co-production activities of public service organisations-for example, where recovering drug users volunteer to act as peer mentors to current addicts. Finally, by directly involving citizens in how public services are conceived, planned and delivered, co-production may be especially effective in bringing about behaviour change, as direct experience may be a more convincing trigger to putting their life right, as compared to being subjected to publicity campaigns or having their choices framed for them in controlled 'nudges'e.g. where peer mentors gain so much self-confidence and self-esteem that they entirely kick their drug habit and raise their ambitions to achieve challenging qualifications and more rewarding employment.

The Public Value Model in Fig. 2 .1 goes still further-it highlights the importance that resilience plays in creating and maintaining public value, since co-production affects the capacity of a stakeholder (either commissioning and provider organisations or service users, communities and other stakeholders in the overall system) to cope with disruption to their current quality-of-life outcomes. This is an issue which has been largely neglected in co-production research to date (Bovaird and Quirk 2017)-but see the thought-provoking discussion by Coaffee, Albuquerque and Pitidis (Chapter 28) in this volume.

Finally, the Model highlights that it is essential that public governance principles inform both the way in which services are commissioned and provided, and the ways in which co-production and behaviour change are mobilised.

Here we see a critically important difference between value creation in the public sphere and in private markets: 'the ends do not justify the means'. Co-production must therefore be implemented in full accord with public governance principles, which are likely to include, among others, transparency, accountability, fairness and due process, the equalities agenda and sustainability. Moreover, these principles, and the priorities between them, must be at least partly co-produced with citizens.

Clearly, in co-production contexts, these public governance principles should cover the behaviour of both citizens and public service organisations. Accountability should mean that each should be able to hold each other to account for how well they have performed their agreed contributions. Similarly, transparency should mean that both professional staff and citizens should be open with each other about what they are up to in their co-produced activities, both sets of co-producers should act fairly towards each other, and co-production practices should involve behaviour which is sustainable by both parties. Through this relationship of reciprocity, co-production may make it more likely that these public governance principles will be adhered to by the co-producers themselves. Indeed, the achievement of these public governance principles can be seen as a separate set of outcomes to which co-production can make an important contribution.

Co-production therefore plays four major roles in this Public Value Model:

• co-production can enable services commissioned by the public sector to achieve higher levels of quality-of-life outcomes through making better use of the contributions of citizens (either service users or community members); • co-production can enable publicly desired outcomes to be achieved by citizen co-producers who help public service organisations but without the need for formal public services (e.g. community members informally checking on and chatting with elderly or infirm neighbours as a result of public campaigns for more solidarity during the COVID-19 crisis, without having to register with public service organisations as formal volunteers); • co-production can strengthen the resilience of the outcomes of the overall public value system and each of its component parts (e.g. resilience of users, resilience of their communities, resilience of their individual provider organisations and resilience of the overall set of providers); • co-production can help to formulate public governance principles, decide the priority between them, and ensure that they are adhered to in the public realm-and co-production must itself be practised in accordance with these public governance principles.

At the same time as these major roles, we should recognise that co-production may enter into all of the other elements of Fig. 2 .1 to some extent. For example, although directly elected politicians may have final responsibility for determining publicly recognised 'needs', they are likely to seek and take into account the views of citizens in their decisions-indeed, this is often a key element of the co-commissioning of public services, as we will see later in this chapter. Again, the operation of public service organisations may be partly influenced by citizen representatives on their boards (although this is typically much more important in third sector organisations, rather than in public or private sector organisations). And how service users, communities or businesses experience and assess their own outcomes does not emerge simply from their own interior reflections but is at least partly influenced, in a co-productive relationship, by the interactions which these stakeholders have with people in public service organisations, such as teachers, doctors, social workers, etc. Finally, this is a Public Value model and does not purport to represent the full range of ways in which civil society creates value through individual self-help and the self-organisation of communities-aspects of life which may be responsible for much higher levels of value creation than either the public domain or the market system. However, there is clearly a strong two-way relationship between this public value model and value creation in civil society-successful self-help and self-organisation reduce the need for public sector interventions and vice versa. Moreover, skills developed in co-production of public value may contribute substantially to self-help and self-organisation and vice versa. This chapter focuses on co-production but it will always be important to keep in mind these interrelationships with purely civil society modes of value creation. Finally, we need to recognise that the Public Value Model in Fig. 2 .1 is very likely to represent a complex adaptive system. This is evident not only because the model is circular, as opposed to being simply linear, but also because there are so many linkages between elements within the model itself. This has important consequences for the planning, implementation and evaluation of co-production initiatives. The basic lesson from complexity sciences is that 'context matters', because it is co-created at the same time as the outcome targets and the policy levers which public policymakers put in place (Bovaird 2013) . Putting substantial faith in clear cause-and-effect chains, either in setting strategy or in evaluating policy, may work well in some parts of the service system, where we are working within simple or even 'complicated' knowledge domains but is entirely inappropriate in the 'complex' (never mind the 'chaotic') knowledge domain. Here, we cannot predict likely positive or negative co-production outcomes of the interventions by either professionals or citizens-the best we can do is to experiment and sense what works and what doesn't (Loeffler 2021) . Of course, even in complex adaptive systems there remains a role for meta-planning-however, far from predicting the direct consequences of co-production interventions, this must limit itself to attempting to influence the envelope of feasible system outcomes by shaping the underlying parameters of the system. Public policy evaluation has long taught us that clarity of long-term impacts is difficult to operationalise even in the best of circumstances but it is not even a sensible goal in the complex knowledge domain (Bovaird 2012) .

What does this mean for co-production? It may well be that many coproduction initiatives are undertaken in the simple or complicated knowledge domains, so that linear models are applicable. However, the very nature of co-production, as highlighted in Fig. 2 .1, suggests that it often takes place within complex adaptive systems. As Jon Pierre (2013) suggests, such situations, where context is so important, may mean that generalization, and therefore 'cumulative science', is very difficult. However, the corollary may be that some aspects of public policy and management will remain relatively impervious to 'science' in the narrow sense, and be better dealt with by craft or art-and this should not be regarded as a defeat for 'knowledge' but rather a recognition of the need to contextualise what kinds of knowledge are relevant in tackling different classes of problem (Bovaird 2013 ).

Co-production as Citizen Voice and Citizen Action: The Four Co's As Table 2 .1 highlights, many of the early writers on co-production focused on the role of citizens in the co-delivery of public services. Part of their intent was clearly to point out how citizen contributions to the public domain were not simply confined to public consultation and participation, topics which had been of great interest in public administration since the 1960s. However, it was inevitable that this attention to 'citizen action' would eventually be aligned with the preceding interest in 'citizen voice'. This was not only because citizens themselves saw their contributions as holistic but also because the analysis of citizens' role in co-delivery very quickly threw up that citizens were often involved in co-design of the services which they were co-delivering. Co-design, in turn, often followed from complaints about services, which also sometimes stimulated discussions of alternative forms of public interventions to achieve outcomes which current services could not achieve cost-effectively.

Consequently, there have many suggestions over the past decades as to how the co-production concept could be disaggregated into different approaches to co-production-for example, Bovaird and Loeffler (2012) listed co-planning, co-design, co-prioritisation, co-financing, co-managing, co-delivery and co-assessment (including co-monitoring and co-evaluation). Others quickly added in co-decision, co-governance, co-implementation, co-construction, etc., leading to Coaffee and colleagues (in this volume) describing these approaches generically as 'co-ubiquity'.

In order to simplify, we suggested in 2013 that most of these various terms, each valuable for specific purposes, can usefully be aggregated into the Four Co's-co-commissioning, co-design, co-delivery and co-assessment (Bovaird and Loeffler 2013a):

• Co-commissioning of services and outcomes, which embraces:

-Co-planning of policy-e.g. deliberative participation, citizens assemblies, -Co-prioritisation of services-e.g. participatory budgeting (macroscale), personal budgets (micro-scale), stakeholder groups on budget committees. -Co-financing of services-e.g. crowdfunding, local fundraising.

• Co-design of services-e.g. user forums, service design labs.

• Co-delivery of services and outcomes, which embraces:

-Co-management of services-e.g. community management of public assets (such as libraries, community centres, youth clubs, sports facilities), school governors. -Co-implementation of services-e.g. peer support groups, Neighbourhood Watch, Speed Watch).

• Co-assessment (including co-monitoring and co-evaluation) of services and outcomes-e.g. tenant inspectors, user online ratings.

As well as simplifying what can otherwise be a confused and confusing narrative, distinguishing the Four Co's in this way is in line with the spirit of the recommendation by Brandsen and Honingh (2016, 433 ) that 'we must depart from a single usage of the term 'co-production' and start using conceptually more distinct varieties'. In the interpretation of the Four Co's, it is important to remember our argument that not all public sector contributions to public value are through the medium of public services-e.g. tackling inequality is a common policy objective but its achievement requires a raft of policies, strategies, regulatory mechanisms, taxation and welfare benefits, as well as public services. Consequently, we need to see the Four Co's as being about the achievement of outcomes, as well as services.

Of course, this framework has its limitation. The boundaries between the Four Co's are not watertight-for example, co-commissioning of which services are high priority can depend on how those services are designed, so both activities need to be done simultaneously. Similarly, the practical codesign of a service may benefit greatly from experience of co-delivery, so these two elements of co-production may valuably be intertwined. Again, Strokosch and Osborne, in this volume, define 'co-construction' as the two-way process of public service delivery, by which service users, their networks and service providers act jointly to frame the value and outcomes that they can derive from service delivery (and, indeed, from their whole-life experience), which is an intertwining of co-delivery and co-commissioning.

Moreover, not all the 'co-ubiquitous' words fit neatly into this framework. The phrase 'co-decide' is quite common and clearly could apply to any of the Four Co's. The phrase 'co-governance', in so far as it is applied to citizens (rather than third sector organisations) typically has three quite distinct meanings-one is an element of co-commissioning, namely where citizens help set to set the strategy of public service organisations by prioritising goals, outcomes, services, etc. The second is an element of co-management, where citizens play a role in the management of public service organisations or public services. Third, it describes the co-assessment role of citizens in ensuring that public governance principles are adhered to by public service organisations.

However, the distinction between the Four Co's is not simply an academic exercise-it has important real-world consequences. First, three of the Four Co's are about citizen voice-co-commissioning, co-design and coassessment-while co-delivery is about citizen action. Generally, the people who get involved are quite different between these two kinds of coproduction-citizens who are most often involved in citizen action are often relatively uninterested in the typical manifestations of citizen voice-meetings, committees, writing letters, making complaints, etc. And, similarly, people who are highly effective at making their voice heard may be rather unwillingindeed, they may feel they simply have not got the time-to put in the hours needed to patrol in StreetWatch, pick up litter in public parks, visit and comfort local neighbours and attend fires as volunteer fire fighters. One of the main reasons for respecting the advice from Brandsen and Honingh (2016, 433) to 'start using more conceptual varieties' of co-production is to ensure that recruitment of citizens for co-production takes into account for which of the Four Co's a particular citizen has appropriate strengths and interests.

A further practical consequence of this distinction between citizen voice and citizen action is the need to reconceptualise the role of citizens in democracy. This is often discussed in terms of citizens' rights and responsibilities in decision-making, i.e. in using their voice to effect change. However, coproduction through citizen action may be at least as powerful a way of effecting change in society by making citizens' beliefs, values and priorities evident and then realising them through social activities. Co-production of the people, by the people and for the people may have a governance significance which is much wider than simply its effect on public services and outcomes.

Change Management of Co-production:

The Co-production Star Model

One important theme which has been given little attention in the coproduction literature is the change management process which can expedite the move to greater and more intensive co-production on the part of both citizens and members of public service organisations. In this section, we highlight one change management tool which has been used successfully in practical settings to develop, test and scale co-production approaches-the Governance International Co-Production Star (Fig. 2.2 ). In the inner ring of this figure are the Four Co's of co-production, while the outer area has the five change management steps-mapping existing co-production initiatives, focussing on those with the highest impact, involving the right people, inside and outside the organisation, who can make the strategy succeed, marketing it to the sceptics and growing it within and beyond the organisation (Loeffler and Hine-Hughes 2013).

Step 1: Map It! Most public service organisations are already taking some steps to increase the effectiveness of at least one of the Four Co's. However, they are often only dimly aware of how much their staff are already doing, have recently tried or are wishing to experiment with co-production approaches. Consequently, mapping recent, current and potential future co-production initiatives is key to making sensible decisions on the direction in which to take coproduction. This mapping will be most powerful when it blends creativity with existing knowledge, so it needs to bring in the experience of a wide range of stakeholders, but especially frontline staff, service users who are 'experts by experience' and people with knowledge of innovative practices from national and international case studies. Such mapping exercises will also be more informative when they separate the four modes of co-productionco-commissioning, co-design, co-delivery and co-assessment-so that future ways forward under each mode can be considered.

Mapping what is already being done and what potential future co-production initiatives might be undertaken provides a list of options for developing the next steps. Prioritisation is then essential-the issue is how to focus strategically on the areas where co-production is likely to work best and be the most cost-effective way of achieving outcomes. This clearly needs to be done separately for each of the Four Co's. While there will often be 'quick wins', the case for which is obvious, there is likely to be a much longer list of co-production initiatives which are promising but uncertain in the outcome and efficiency gains which they might bring. In selecting the options to take forward, an outline business case for each is likely to be helpful, especially for those co-production activities which involve significant spend or which mean a major change in direction in a service. However, given what we have said above about working in a complex knowledge domain, it will be important to undertake a range of creative options, even though they appear to have a significant risk of failure, since what works and what doesn't work will more likely emerge from experience and experimentation than from prior planning.

The third step is to determine who are going to be the co-producers in the initial piloting of each co-production initiative which has been prioritised in step 2. It will be important to have committed, motivated and skilled individuals to take these experiments forward in a creative way.

In order to identify who can make which contributions, it is helpful to undertake 'capabilities assessments' with all involved co-producers, which can complement the more common 'needs assessment' process in public services. These identify the strengths, assets and potential contributions that co-producers might make to improving their own outcomes and those of other citizens.

This information helps to identify the citizens and staff who are either actual or potential co-producers, and who can then work together in innovative and practical co-production initiatives in which they themselves want to be engaged. There are various ways in which experimentation can occur but our experience has been that 'Co-production Labs', based on design principles, are especially effective in generating 'buy-in', not only from citizens but also from staff members (Loeffler and Schulze-Böing 2020) . Other stakeholders, too, can be valuable participants.

Co-production can only work if the stakeholders involved are committed to making it successful. Although there is a significant literature on what motivates citizens and staff to co-produce (see Chapters 25 and 26 in this volume), there is much less research on policy mechanisms which can activate these motivations in practice, so keeping current co-producers on board and attracting new people who want to join in.

A key element of such mechanisms is likely to be providing evidence of how successful co-production can be in appropriate circumstances. This can be partly achieved through formal evaluation approaches (Loeffler and Bovaird 2018) but, in complex adaptive systems, may be more convincing when undertaken through rapid appraisal to weed out non-effective options and highlight those options likely to be most cost-effective. Of course, when the number of options has been reduced, more thorough evaluation becomes more desirable, if feasible. However, even when substantial evidence is available that co-production works (or, in some cases, that it doesn't), it is unrealistic to expect that policymakers will be convinced-as Tinna Nielsen points out in Chapter 5, decisions are often made on non-rational grounds rather than on evidence, so these non-rational drivers also need to be emphasised when marketing the potential role of co-production.

Step 5: Grow It! Experimenting with and demonstrating the cost-effectiveness of specific coproduction initiatives is only the first part of the story. Scaling up successful co-production (and stopping ineffective co-production) requires the willingness, both in public service organisations and in civil society, to change old cultures, beliefs and habits. This means that the co-producers (both citizens and public service organisations) have to exhibit the positive characteristics of a willingness to change (adaptability, innovativeness, entrepreneurialism, risk-accepting) and at the same time have the capabilities to overcome the barriers to co-production (organisational, contextual, user-characteristicsderived, community-characteristics-derived) . In Fig. 2 .3 we set out in more detail a conceptual model for understanding this set of barriers. Clearly, in order for these barriers to be overcome, a range of strategies will be needed on the part of all the stakeholders involved-this will involve user strategies to tackle the user-characteristics-derived barriers (e.g. a developmental strategy to improve users' skills needed for co-production), community strategies to tackle the community-characteristics-derived barriers (e.g. a digital networking strategy to increase social connectedness in the community) and organisational strategies to overcome organisational barriers (e.g. a risk enablement strategy in order for public service organisations to become open to more promising opportunities for co-production).

In the process of growing the co-production approach, the role of citizens is likely to be critically important. Community champions who have been 'early adopters' can help to inspire and mobilise other members of their communities, e.g. by making presentations in co-production roadshows, writing blogs, and visiting community groups to activate more participants.

Case Study: Implementing the Co-production Star in Jobcenter MainArbeit, Germany

Jobcenter MainArbeit in the State of Hesse has had a long-standing reputation in Germany as a particularly innovative employment agency. It is a municipal agency (Eigenbetrieb), managed and governed by the City of Offenbach, near Frankfurt am Main, with about 125,000 inhabitants, including a high proportion of migrants.

In 2019 a core group of about 30 staff members of Jobcenter Mainarbeit and representatives of third sector organisations embarked on a co-production journey, based on the Co-production Star (Loeffler and Schulze-Böing 2020). The first workshop enabled participants to shape new co-production initiatives, based on extensive mapping of current co-production and ideas for potential new initiatives. Participants were surprised to identify that many of their activities already involved elements of co-production-but they also recognised that, apart from a few exceptions, there was underused potential to strengthen co-production in many counselling and job placement activities. This raised the question of where to start? The second Co-Production Star Workshop involved an iterative prioritisation process, enabling participants to reflect on which co-production initiatives were most likely to improve key outcomesand directly stimulating a debate on which quality of life outcomes were most important for Jobcenter Mainarbeit.

The third Co-production Star Workshop prepared participants for the Co-Production Labs by introducing them to the methodology of the Governance International Co-design toolkit, identifying potential risks and providing them with space to discuss how to approach and invite service users to join each specific Co-Production Lab. Participation of service users was agreed to be completely voluntary. Furthermore, the Chief Executive of the Jobcenter made it clear that the Labs were free to experiment, even it this led to 'mistakes', as long as they learnt from them and shared the lessons. Governance International supported the Labs during the 100 days of experimentation (which were extended to 120 days as they started during the summer holiday period) with facilitation of meetings with service users and workbooks to document the lessons learnt. Altogether five Co-Production Labs were launched involving the following co-production initiatives:

• Co-commissioning 'personal employability budgets' with young jobseekers. • Co-designing a 'Citizen Symposium' with service users getting ready for retirement. • Co-designing service offers from Luise 34, a social enterprise for the unemployed. • Co-delivering peer support for and with people seeking a job.

• Co-assessing training courses and other projects with participants.

The Co-production Labs progressed at different speeds but all managed to test some small-scale actions and to share the learning with the other Labs.

The fourth Workshop provided the Co-Production Lab teams with tools to develop pathways to outcomes and assess changes to quality of life outcomes, both quantitatively and qualitatively. In the fifth Workshop the participants identified barriers to co-production within and beyond their organisation and discussed strategies for overcoming them in order to scale their successful coproduction initiatives. The teams also discussed the limits to co-production in employment services and agreed that not all services provided by the Jobcenter Mainarbeit and its partners would be suitable for co-production.

The five Co-Production Lab teams presented the lessons learnt and initial evidence of the results of their Lab at a Co-production Celebration Event at the end of the year, which provided all co-producers with the opportunity to value the contributions everybody had made and to reflect together on the next steps. While some Co-Production Labs considered that a redesign of their initial co-production strategy was necessary, other Labs wished to scale their prototype projects and work with other groups-this was particularly the case with the Lab developing a peer support mechanism for young job seekers (Neseli and Herpich 2020) , where the initial experiment proved to be very successful, motivating staff to extend this approach to other areas of work in Jobcenter Mainarbeit.

Co-production is, above all, a collaborative practice. When this practice is formally approved by a public service organisation, it acquires the status of a policy or a strategy. And when it becomes the accepted and expected way of doing things-which is not yet very common in the public domain but is sometimes the case in community development or neighbourhood management-it can be considered an institution. It is generally voluntary, i.e. not legally mandated nor regulated-although some authors (e.g. Bertelli and Cannas in this volume) argue that in some circumstances it could indeed become mandatory. We take the stance that such coercion is only likely to work if it is short-term, leading co-producers to recognise the value of the collaborative actions, after which they then can join in voluntarily.

Co-production is not an ideology, although it is given very strong backing by adherents of some ideologies, especially those which give major weight to participative democracy, subsidiarity and decentralised power. Whether or not it is a normative concept is also an issue on which authors in the literature (and this Handbook) differ. Some do appear to regard co-production as a 'good' practice, always to be encouraged. However, most caution that there can be a downside to co-production (see especially Smith in Chapter 31 of this Handbook). We take the stance that co-production, as a concept, is about a practice which can have both positive and negative sides, so needs to be examined critically.

Although there are already many contending definitions of co-production, they still have a common core of elements which means that user and community co-production of public services and outcomes remains a relatively coherent theme in the social sciences literature. Moreover, most definitions contain elements of both citizen action and citizen voice, and most approaches to co-production can be categorised as falling within the range of the Four Co's-co-commissioning, co-design, co-delivery and co-assessment.

Co-production plays a key role in creating public value-it can help to improve publicly desired outcomes directly, to make public services more costeffective, to strengthen whole-system resilience and to support the achievement of public governance principles. By the same token, its absence or its misapplication can reduce public value.

Understanding of the change management necessary to operationalise greater and more cost-effective co-production remains relatively underdeveloped. In this chapter, we have suggested a five-step change management process which has recently proved valuable in some practical settings. However, further research on this would seem an urgent priority.

Engaging Public Sector Clients: From Service-Delivery to Coproduction

Beyond Engagement and Participation-User and Community Co-production of Public Services

Attributing Outcomes to Social Policy Interventions-'Gold Standard' or 'Fool's Gold' in Public Policy and Management? Social Policy and Administration

Context in Public Policy: Implications of Complexity Theory

From Engagement to Co-production: How Users and Communities Contribute to Public Services

We're All in This Together: Harnessing User and Community Co-production of Public Outcomes

The Role of Co-production for Better Health and Well-Being: Why We Need to Change

The New Commissioning Model of Services for Young People in Surrey: Evaluation of Achievements and Implications

Resilience in Public Administration: Moving from Risk Avoidance to Assuring Public Policy Outcomes Through the Resilience Chain

The Challenge of Co-production: How Equal Partnerships Between Professionals and the Public Are Crucial to Improving Public Services

Distinguishing Different Types of Coproduction: A Conceptual Analysis Based on the Classical Definitions

Toward a Definition of the Coproduction Concept

Representative Co-production: Broadening the Scope of the Public Service Logic

Co-production of Public Services and Outcomes

User and Community Co-Production of Public Services: What Does the Evidence Tell Us?

Assessing the Effect of Co-production on Outcomes, Service Quality and Efficiency

Co-commissioning of Public Services and Outcomes in the UK: Bringing Co-Production into the Strategic Commissioning Cycle

Five Steps to Making the Transformation to Co-production

Co-production Star Action Learning Programme in the Offenbach Employment Agency: Results and Lessons from the Five Co-production Labs

Creating Public Value: Strategic Management in Government

Job Seekers Providing Peer Support for Each Other in the Offenbach Job Agency. Governance International Case Study

Crossing the Great Divide: Coproduction, Synergy and Development

Consumers as Co-producers of Public Services: Some Economic and Institutional Considerations

Citizen Participation in the Coproduction of Urban Services

Context, Theory and Rationality: An Uneasy Relationship?

A Systematic Review of Co-creation and Co-production: Embarking on the Social Innovation Journey