Public harm or public value? Towards coproduction in research with communities

This paper develops a critique of the current model of research governance ethics which casts communities as vulnerable subjects. The paper constructs an alternative approach to thinking about the twin challenges of research governance and reflexive research practice through reframing ideas of public value and rejecting a public harm model. We use the insights of coproduction as a way of positively rethinking the relationship between researchers and ‘the researched’ to create new ways of thinking about public value. We argue that reflexive research practice should draw upon the principles and examples of empowerment within governance. These aspire to work ‘with’ communities and create space for greater community control in defining and creating publically valuable research.


Introduction
Higher education institutions in the UK have long faced an accusation of a 'relevance gap' in research (British Academy, 2008).The challenge of 'relevance' may be broadly considered, referring to the potential contribution of universities to local and regional economic development (Chatterton, 2000).More recently, there has been a specific return to the 'civic' role of universities, implying commitment to greater engagement with the local communities they are situated in and serve (Goddard, 2009).This challenge overlaps with the demand for 'impact' made in the Research Excellence Framework (REF), the revised system of assessing research quality within higher education (Armstrong and Alsop, 2010), which calls for publicly funded research to evidence its social value and importance.Some have presented scathing opinions on such attempts to audit 'relevance' (Slater, 2011) and have asserted that REF measures continue a neoliberal emphasis on individual accountability and disincentivise collaborative forms of research practice (Dowling, 2008).Others reflect that the REF challenges those researchers who write about social injustices but lack explicit linkages to effecting social change and propose overcoming limitations in current systems of research governance by giving attention to the processes as well as the outputs of research (Pain et al, 2011).
The last few decades have seen a cross-disciplinary proliferation of research that aspires to encourage greater community participation and empowerment within decision making (Arnstein, 1969;Barnes, 2009).Evidence of the depth and richness of such research is found in critical geography (Pain, 2003;2004), development studies (Gaventa, 2005), sociology (Burawoy, 2005), and anthropology (Falzone, 2004).Such works highlight participatory approaches as highly normative with a value base in democracy, social justice, and human rights (Fishkin, 2010;Heron, 1971;Heron and Reason, 1997;2001;Lewin, 1946;Narayan et al, 1999).As Pain et al highlight, "Participatory approaches cannot circumvent the paradoxes of power in research and representation … but through closer integration of theory and practice, they extend the processes of theorising and knowing beyond campus spaces" (2012, page 121).Such approaches aspire to move from the merely 'dialogical', where research seeks to establish some form of conversation between the researchers and the researched, towards 'transformative research' where the purpose is to "both engage the researched at the problem definition stage and to actively alter the social conditions in which they find themselves" (Robinson and Tansey, 2006, page 152).
We concur with the challenges made within participatory geographies to the wider ethos of the research process, but this aspiration presents twin challenges: the current context of research governance, as suggested in the contradictory responses to the REF, and the current space for reflexive research practice.In this paper we discuss how ideas of 'public value' provide an alternative set of principles for research governance, which currently has its basis in a bureaucratic and medical model of minimising harm towards communities.Kesby argues that we need to expand our notion of ethical research beyond the "conventional wisdom of 'do no harm'; by crafting new spaces for critical engagement … participants and researchers can collaboratively generate knowledge and informed action" (2007, page 2813).We also explore the potential of reflexive practice and the importance of the reflexive researcher to provide alternative understandings of ethical responsibilities and the relationship between the researcher and 'the researched'.Coproduction builds on work in the field of participatory action research (PAR) which suggests how to "design research to reveal spots of possibility, extraordinary spaces where democratic practice could/does take place" (Fine and Torre, 2004, page 29).
We use the concept of 'public value' (Moore, 1996) and the ethos of coproduction to reimagine the current model of research governance.This way of thinking about public value (Moore, 1996) has emerged as a means of rebuilding legitimacy for public institutions by restating their distinct value; and recognising that public institutions now "need to win, consent, persuade, explain, share responsibility" (Goss, 2001, page 163).Moore (1996) called for a more rounded accountability which recognises that public value is created in a political decision-making context where institutions have responsibilities to communities as citizens.In turn, communities bring knowledge and perspectives that institutions may not possess (Barnes, 2009).
Public value, in this context, differs greatly from prevailing notions of accountability within higher education, which reinforce traditional dynamics of hierarchy and expertise.For Walker (2010), academic practices can be a refuge from engagement, with peer review acting as a form of 'epistemological protectionism' asserting monopoly rights to defining and legitimating knowledge.Bond and Paterson (2005) argued that UK academics adhere to a strong cultural norm that values international peers and audiences rather than local or community concerns.Public value, however, conceives of a more 'nested' notion of accountability which acknowledges the variety of potential action situations (Hupe and Hill, 2007).Accountability is seen as a "social relationship in which an actor feels an obligation to explain and to justify his conduct to some significant other" (Bovens, 1998, page 172).Our accountability to communities which may be research participants or the subject of research findings is becoming more narrowly defined within the realm of institutional regulation and control.Whilst research ethics undertakes a duty of care towards potential participants, accountability is tightly defined.Our current emphasis on 'public harm' sees individuals as research subjects engaging in a discrete section of a research projectas 'informants' to that research.There has been an inexorable rise in the mechanisms to measure impact; but there are few, if any, mechanisms to consider the cumulative influence of research projects for specific communities.Given these contradictions, we consider this growing area of research governance through research ethics oversight to question how it is currently framed and to propose ways in which it may be framed more effectively with communities.
Criticism of institutional forms of knowledge and action is not new.Freire (1970) felt it was critical for there to be counterhegemonic approaches to knowledge construction in oppressed communities to challenge the dominance of majority or more powerful interests and perspectives.The Autonomous Geographies Collective's honest account of an action research project reminds us that "participatory research is not inherently progressive" (emphasis in original) but rather needs to draw upon activist approaches to signpost key strategies for ethically concerned research (2010, page 249).Though some commentators have emphasised that coproduction concerns itself more with the construction of knowledge than progressive social change (Gillard et al, 2012), others have argued that the two are importantly related.To realise ambitions of coproduction, research should go further and challenge the embedded knowledge hierarchies of the expert versus the layperson (Porter, 2010).As Pohl et al acknowledge, "co-production of knowledge requires that contributions from specific disciplines and social actors are not privileged over what other disciplines and social actors contribute"; "it also requires that communication is not seen as a oneway transfer from a knowing subject to a supposedly ignorant one" (2010, page 271).For example, Gaventa (2005) asserts that participation in knowledge creation is part of the wider citizenship rights accorded by society, including the right to social justice.
Coproduction emerged in the context of a longer dissatisfaction with participation which has been accused of informing or manipulating communities but not transferring power to them (Arnstein, 1969, Cooke andKothari, 2001).Kesby (2007Kesby ( , page 2827) is critical of simplistic accounts of 'participation as tyranny' (Cooke and Kothari, 2001).Whilst he acknowledges that such practices are 'power ridden', he argues that participation provides meaningful and important opportunities to 'facilitate change' in the lives of others, if we are self-reflexive and cognisant of these entanglements.Coproduction provides a useful heuristic to encourage reflexivity (Beebeejaun et al, 2013;Durose et al, 2012;Pain et al, 2011;Slater, 2011).The inclusion of other perspectives throughout allows exploration of how "respective contributions to the research process were socially situated and what this means for the knowledge that was produced" (Gillard et al, 2012(Gillard et al, , page 1135)).Whilst coproduction may not be appropriate for every research study, there is a need to critically examine the conditions in which coproductive research can flourish if it is to have wider resonance in research and in creating fuller definitions of public value.
The focus on potential public harm to communities infuses institutional research ethics and governance, and this is the concern of our paper.We propose an alternative focus in institutional research governance away from the paradigm of public harm which restricts community voice toward a way of thinking centred upon public value.We suggest that reconsidering the concept of 'public value' unlocks coproduction as a means of engaging substantively with communities.We now consider in greater detail deficiencies within current university research governance.
Reimagining research governance: from public harm to public value?

Reflecting on the limitations of a public harm model of research governance
Universities have sought to formalise policies and procedures for ethical review, drawing on a 'medical model' where institutional ethical review precedes community involvement and placing ethics committees as safeguards between the researcher and the research participants (Dyer and Demeritt, 2009).The rationale of a medical model of research is clear: research intervenes directly on the physical well-being of the subject and carries an inherent health risk, and the role of research governance and ethics is to avoid 'public harm'.The role of ethical review is premised on reducing the potential harm that may occur to participants and the potential liability to institutions.This process is appropriate for medical research or other research which potentially has a direct physical effect on the participant-for example, through clinical trials or tests where monitoring research can be used to minimise individual harm.In this current framework potential researchers must be explicit about the topics of questions, the recruitment process for participants, the time length of interviews, and the safety of the data.One illustration of perceived public harm is a question from an ethics committee application form: "What are the potential adverse effects, risks or hazards for research participants, including potential for pain, discomfort, distress, inconvenience or changes to lifestyle for research participants?"(University of X, no date, page 6).Answers from qualitative researchers submitting to this committee have included tiredness or possible discomfort from sitting in a chair for too long.Ethical approval for detailed research design implies research presents risks whoever participates and however it is conducted.Dyer and Demeritt (2009) question the appropriateness of a research model in the sociological and humanities disciplines where the risks are more amorphous.We question whether risk should be understood through potential harm or inconvenience to individuals.Whilst well-designed research that shows care towards people is essential, this idea of reducing risk to individuals may not be as helpful as it appears.Dyer and Demeritt highlight that socially progressive research may challenge the perspectives of some participants, or attempt to uncover injustices and therefore not safeguard individuals from social risks.Hay (1998) writes that "no inflexible code of ethics can be contemplated seriously" (page 56), but as Haggerty (2004) reflects in his discussion of the Canadian Institutional Review Board (IRB) process, the bureaucratisation of the process has led to extensive rulebooks and guidelines that attempt to anticipate all research possibilities.The codifying of research practices through institutional ethical review process directs researchers to reaffirming a set of boundaries between them and communities.It precludes the development of relationships of trust between parties and flexibility of research practices in the field, and ignores peoples' abilities to make choices or to engage constructively.Attempts to comprehensively outline all potential risks lead to a set of stringent guidelines that make it difficult to become a reflexive researcher who can make changes when necessary and helpful in the field (Buchanan et al, 1988).Whilst debates about professional ethics are essential and live, the reality of research ethics governance is the erosion of space to discuss these issues in a way that has meaningful influence within institutional oversight.
There is also a false mask of neutrality within the establishment of codes, which sustain power asymmetries in our apparent concern with vulnerable communities (Hay, 1998).This characterisation can overemphasise communities as problem-ridden and researchers as experts who will 'save' them.The Durham Community Research Team note that these processes emphasises the researcher-subject hierarchy and may "protect institutional power at the expense of community empowerment" (2012, page 9).Protocols such as gaining signed informed consent after reading plain-English explanations of our research play an important role in ensuring participants are aware of research processes.But they fail to democratise the research process and can create further distance between researchers and communities who are cast as vulnerable subjects.This closes down opportunities for research participants to move beyond their role as subjects and engage with the researcher or institution in different ways.It reinforces notions of power and hierarchy between researchers and communities that coproductive research seeks to challenge.There is a leap of faith that community members must take in trusting the researcher to accurately represent their words, and contribute to the goals or values of their community.Though university ethics require informed consent, the only option for informants is to participate or withdraw in many studies.Practically, the level of accountability to communities is constrained and limited.The principal right a participant has is to deny usage of his or her interview data.Placing the participant as subject removes discussions about his or her agency and the reasons for participation.Accountability occurs only in relation to the individual but is not placed within wider networks of accountability to communities or to other ethical considerations, such as progressive practice, which PAR emphasises.
More perniciously, the public harm model prioritises the risk rather than the benefits of research.As each potential interaction is posed as causing potential risks, it makes aims to create and sustain relationships with communities difficult to justify to ethics committees.Potential benefits are framed as immediate and of direct benefit to the individual participant at that time or nonexistent.These institutional norms do not imbue communities with the power to influence their own lives or those of researchers.The higher education sector risks becoming increasingly out of step with political values, which over the last few decades have emphasised citizen power and state responsiveness (Arnstein, 1969).Greater researcher openness in how research studies become framed is not seriously addressed through the current ethics model.
We do not make light of responsible and ethical research practices which display sensitivity towards participants and are undertaken in a spirit that recognises the goodwill displayed towards us when people are willing to participate.Our criticism is that the guidance does not represent either researchers or participants as people who negotiate a social world in everyday life.Decisions about risk remove participants further from the governance process because they are invited into research only at a comparatively late stage and their role is tightly constrained.The potential social risks or consequences of research might be ones that communities are willing to take or support.However, the absence of a space for communities to engage in setting research agendas precludes these discussions.
If the rationale of public value is to be believable from the viewpoint of communities, their role must be recast beyond that of vulnerable subjects.Coproduction as a means of generating public value allows us to reimagine the role of communities in research and challenge the current model which seeks minimal 'intervention' into participants' time but which limits their role to informants to increase our knowledge rather than potential collaborators.It also reevaluates existing bureaucratic processes to question whether there are more productive ways of thinking about research ethics that create spaces for open discussion rather than closed forms of regulation.Rethinking research governance has the potential to create new 'shared thought spaces' (Pohl et al, 2010) and realise the transformative ambitions of research (Robinson and Tansey, 2006).We now consider the potential of coproduction in unpacking institutional research ethics through linking it to governance practices which seek to collaborate with communities.

Towards a mode of public value: learning from governance in communities and internationally
The rise of a paradigm of risk minimisation in research ethics governance creates distance between researchers and communities.Here it is useful to consider, in contrast, how engaging with citizens and aspiring to empower communities has become increasingly important in wider governance practices.Successive UK government programmes have attempted to intervene and improve the life of communities in deprived areas.From the Urban Programme onwards, differing levels of community engagement have been proposed as a means of creating ownership and better policy making but remain hampered by state agendas (Taylor, 2007).The proliferation of the notion of governance in theoretical and practice-led debates can be misleading; it has been argued that the idea of governance's widespread acceptance is oversimplified (Newman, 2005).We must remember that governance is not an agreed set of practices, outlooks, or perspectives, as the "extent to which different conceptions of neighbourhood governance are based upon different democratic values and principles" (Lowndes and Sullivan, 2008, page 71).
For example, New Deal for Communities (NDC), the UK's most extensive experiment in community participation in governance, was premised on the aspirations of community control within the partnership boards.This initiative theorised that community ownership would ameliorate the past mistakes of top-down regeneration programmes (Durose and Rees, 2012).Alongside others, the authors have been involved in critiques of NDC practices as limited in their ability to enable communities to become truly empowered (see also, Lawless, 2011;Lowndes and Sullivan, 2008;Taylor, 2007).Taylor acknowledges spaces were created for community influence, but "new governance spaces are still inscribed with a state agenda, with responsibilities pushed down to communities and individuals, at the same time that control is retained by the centre …" (2007, page 314).Whatever the shortcomings in practice, it provided opportunity to try alternate forms of governance and learn from experience.It is surprising there has not been more reflection on this, now extensive, literature in conjunction with PAR or coproduction debates.
The governance literature demonstrates that the reasons why people make choices to give of their time are complex and multifaceted.Their expectations of the impact of their actions may also be altered during the process.Alford recognises that citizens "donate their valuable time and effort to the achievement of organisational or program purposes … when they receive, or expect to receive, something at least as valuable in return …" (2009, page 188, emphasis in the original).A principal driver for participation is concern about problems facing their communities and a motivation to contribute to positive social change (Beebeejaun and Grimshaw, 2011).Dinham speaks of NDC-elected representations feeling "a great sense of responsibility" (2005, page 308).People felt obliged to participate; and as Barnes reminds us, "knowledge and expertise are not the sole preserve of officials" (2009, page 47).Harnessing people's concerns for wider social issues also chimes with public value agendas (Moore, 1996).Though some of the outcomes of these programmes have been flawed, they offer directions in which to consider the role of communities in providing meaningful and effective input into debates about social problems.Their alternate perspectives can be of value in framing research agendas and in understanding the possible unrealised implications for particular communities before research occurs.The most important aspiration for citizen control was demonstrated through a majority of the governing board representatives coming from the community.In contrast, research ethics boards are, on the whole, drawn from university staff.
A crucial insight from work on governance and specifically deliberative democracy is that a politics of presence counters power imbalances by representing minority interests, and also more fundamentally transforms the nature of research-its framing, processes, and dialogues.Pitkin's (1967) writings outline different forms of representation in political and decision-making spheres; her ideas about symbolic and descriptive forms of representation (rather than formalistic or substantive representation) lay the ground for the idea of the politics of presence (Phillips, 1995).The presence of marginalised groups means that their interests and perspectives either are physically represented (descriptive representation) or are advocated for by those they consider to be their peers, or with whom they share identities or experiences (for example, in certain types of symbolic representation).Young's work on democratic institutional change emphasises the inclusion of underrepresented voices to "correct biases" (2000, page 116).She does not suggest that these groups have more unbiased perspectives but that "the standpoint of those in less privileged positions can reveal otherwise unnoticed bias and partiality" (2000, page 117).
Understanding the skills and knowledge of communities and attempting to address power relationships and imbalances has been a key feature of the shift to governance.Although the outcomes have been mixed, the principles are useful in thinking about research governance.Pouliot, writing in the context of socioscientific issues, highlights that this 'deficit model' creates "a dual divide between citizens and scientists concerning the right to express one's view and the role they are to assume in the production of legitimate knowledge" (2009, page 52).Here, definitions of community can be complicit in conceptualising communities as 'vulnerable'.Often, those living within communities of research are located in areas with complex social problems, or form part of a group that is defined as 'marginalised' or 'hard to reach'.In line with governance literature, researchers must exercise caution with communities which may be wary of state intervention and the purposes to which we might seek to use their knowledge."Local communities have tended to be constructed in an instrumental way by policy-makers: spatially fixed homogenous agents of change with a shared experience of poor housing and social exclusion" (Maginn, 2007, page 28, emphasis in original).Maginn (2007) contends that this instrumental approach shapes how policy makers seek to influence community engagement; labelling communities as deprived or vulnerable suggests they lack skills, and thereby subjugates their knowledge and agency, placing barriers to their involvement in policy making.Public health researchers working with vulnerable communities also have found barriers can be created if they are overtly labelled as vulnerable because it can contribute to social stigmatisation (Dickert and Sugarman, 2005).Thus the framing of research subjects as vulnerable is in opposition to the ethos of community-focused research which looks for the values and assets within communities, which in turn universities may lack.
Evaluations of specific attempts to create community-led and neighbourhood-based governance structures have found that, despite clear goals to have community-led boards, community representatives at times experienced feelings of relative powerlessness (see Beebeejaun and Grimshaw, 2011).Macreavey (2009) shows that, in governance partnerships which include community members, the focus is often on developing community capacity without due recognition that policy makers may well lack crucial skills-this reflects a perception of 'organisational superiority' which alienates and belittles community involvement.One challenge for NDC was harnessing community skills and knowledge through a long process of capacity building and support.Despite the challenges of working in areas suffering from intense deprivation, it is striking that the long-term approach developed capacity in some community members.Dinham found that grassroots community networks emerged and that "there is real growth in people's individual confidence and self-esteem" (2005, page 310).These insights emphasise that current research governance has the potential to alienate communities rather then collaborate with them.
There is also much to be learnt from international examples of community membership in research governance which demonstrate the transformative potential for research practice.Here, community presence is vital in shaping research agendas and priorities at the institutional level.Within US higher education, a number of universities involve community members in partnership boards.These either take an overview on community-focused research across the disciplines or have long-term projects with a particular community.The case of the Detroit Urban Research Centre (URC) which includes academic, professional, and community members provides a thought-provoking example of the benefits and challenges of coproduction within research governance.The partnership " reflect(s) an emphasis on developing and maintaining relationships and infrastructures consistent with community-based participatory research principles, alongside with an emphasis on disseminating findings and knowledge generated" (Lantz et al, 2001, page 496).Detroit URC shows that commitment to coproduction with communities can be effective if it becomes more consistent and embedded within institutional outlooks, engaging with the community to broaden the concept of public value.The URC was aware of the need for scholarly outputs but also innovated through including nonacademic partners as coauthors or copresenters of most papers or presentations (2001, page 501).The Detroit URC provides a framework of broad objectives for research projects and facilitates community involvement in all stages of the research process as well as assisting in dissemination in the appropriate format for the community.
Public value runs through the activities of the partnership, alongside an acknowledgment that priorities for academic and nonacademic partners can conflict, and that the partnership has and will change over time.New forms of partnership working require members to have a reflexive approach that focuses on deriving benefits for the community and researchers over time.Whilst community and research priorities differ, the partnership is evaluated to be a success for both.Such frameworks can support researchers committed to coproduction by providing an institutional setting where coproduction is both recognised and supported.Thus research ethics could engage with wider strands of public value in order to generate a fuller debate with communities and produce outcomes valued by a wider range of audiences.This could include linkages with social responsibility and outreach programmes, to determine broad research priorities, and as a catalyst to stimulate research partnerships.
The preceding examples suggest that community-based research should see participants as more than generators of data for particular research projects and recognise the wider value gained from involving communities within decision making.Coproductive research offers the potential to be successful if it is underpinned by a commitment to opening up to unheard voices and groups.The argument is not about achieving some fiction of balance which replicates traditional models but about the potential to be sensitive to institutional bias.These biases can be revealed and explored through considering research practices and priorities from a range of perspectives including those we research.Within this process, decisions about how to recruit members should be guided by principles around the type of research and knowledge the university seeks to produce.

Coproduction: encouraging the reflexive researcher
Debates which value community presence in research governance can aid in reframing communities as coproducers of research knowledge.Rather than seeing community participants as subjects of academic research, they may offer knowledge, skills, and expertise that can challenge traditional ways of doing things (Fischer, 2000).Coproduction recognises that community members may desire to be more than informers to research produced by others.This is not a simple or easily achievable process.Coproduction rests on developing a shared "thought style" (Pohl et al 2010, page 271).A coproductive research process is importantly premised on creating a 'boundary space' which brings different social worlds-that is, academic research and community life, together.Such spaces need to have "distinct lines of accountability to each" and "involve participation from actors from both sides of the boundary" (Pohl et al 2010, page 268).
UK regeneration initiatives which have aimed to empower communities have demonstrated that there was no one answer to the challenges of governance.One of the strengths has been the process of continuous long-term endeavour which has tested the reflexivity of all participants.The application of lessons learnt through practice has changed the direction of successive programmes (Lawless, 2011;Lawless and Pearson, 2012).Research practices that seek to be coproductive are based on ethical reflexivity and flexible research design and so clash with the rigidity required within institutional ethical review.There is an implicit understanding that communities bring knowledge and ways of thinking that actively transform ways of working and challenge taken-for-granted assumptions within organisations charged with addressing their problems.
Opening up research practices to a wider range of voices requires reflexivity and openness to challenging ways of thinking and working.It presents considerable challenges to all engaged within the process.Action researchers have highlighted the sometimes painful assumptions we have to confront (Autonomous Geographies Collective, 2010).Engaging with communities as active constructers of knowledge introduces a political dynamic into research that is less apparent in traditional research.But there are risks.Achieving public value and relevance for communities is not guaranteed, and failing to achieve expectations may jeopardise future research relations.These challenges are not confined to coproduction but are more keenly felt in such approaches.Coproduction, then, is not so much a solution to our problem but a facilitator in creating the conditions for more relevant solutions.Coproduction usefully contains within it a recognition of the value of transdisciplinary working: that no one discipline is enough, and none has all the relevant knowledge or skills to solve real-world problems.But this recognition presents real challenges to academic concepts of 'expertise' and current measures of research worth.Orr and Bennett (2009) have stressed the need to recognise that in social research there is a dynamic relationship between researchers and subjects, who inevitably influence each other in manifold ways; research is " a social and human practice that embodies institutional, personal and political factors that influence its design, impact and acceptability ... .Reflexivity enables us to highlight the political dynamics of our endeavours, attention to which would otherwise tend to be absent from the representation of our project" (2009, pages 85-87).
What is clear is that such spaces may require facilitation capable of enhancing communicative processes, making different thought styles visible and linking them around common interests.Governance research demonstrates that a commitment to incorporating community voices must also respect their contribution.There is a need to support community members in these roles through a dialogue around current research practice.However, it should not be presumed that new participants are there to adapt to institutional ways of doing things.University participants must also be committed to promoting joint reflection towards a common understanding of situations and collective action-as part of a learning process based on respect, openness, and deliberation (Pohl et al, 2010, page 271).Some participants may span boundaries between the worlds of scholarship and community activism (Newman, 2012).Critical to this blending is an acknowledgement of the validity of lived experiences.Academics must also understand the difficulties to overcome in proposing new ways of working.Stoecker (2008) raises the importance of understanding and working to overcome the practical institutional barriers that militate against community engagement.As mrs kinspaisby reflect in discussing academic roles of teaching, research, administration, and activism, "It's about boundary crossing in many ways" (2008, page 292).Through recognising the interconnectedness of our activities, they highlights the importance of "drawing participatory activities, principles and processes into each of those" (page 292).There has to be considerable commitment to the potential to create new spheres for deliberation, in which the challenging of existing research bias is supported.
One of the most celebrated examples of meaningful community engagement in urban issues is the East St Louis Action Research Project.Established by Ken Reardon in partnership with the University of Illinois at Urbana-Champaign, it evolved from an interest in creating research impact.Researchers received a number of research grants in the late 1980s and decided to find out how effective community representative and public officials considered these programmes.They found either ignorance of the work or belief that these grants benefited the researchers but not the community.What Reardon (1998) terms action research demonstrated the principles of coproduction in involving residents in all stages from research design to analysis.The research grew into a partnership from a number of research projects.Their work ranges from outreach weekends and student projects to developing masterplans or helping to lobby for a metro station (ESLARP, no date).This project has grown over a number of decades and requires sustained dialogue and commitment from all parties as well as considerable funding.It has been successful, leading to a "university empowerment planning approach" that reaped benefits to individuals and the community (1998, page 330).This example highlights the potential for universities to engage with a specific place and the dialogic process which helps to shape research projects and to identify new areas of work.The project contributes to positive physical and social change and guided research and teaching, challenging, and enriching the viewpoint of all participants.The ability to be reflexive was essential, as was courage, which Reardon displayed through honestly asking community members if they had benefited from research activities, which would have clear academic merits.This process of deep engagement has to underpin such research projects in order to reap benefits-a process that can take considerable time and will not always show immediate results.

Conclusions: bringing communities back in
Currently, the public harm model of research governance has, in efforts to protect individuals, led to communities being framed as vulnerable.In contrast to the current focus of the public harm of research, coproduction aims to put principles of public value into practice, working 'with' communities and offering communities greater control over the research process, along with opportunities to learn and reflect from their experience (Bennett and Roberts, 2004).It has transformative potential.Advocates for coproduction argue that research is enhanced through including 'experiential expertise' (Collins and Evans, 2002) which may highlight relevant questions otherwise neglected by 'experts' (Fischer, 2000).
The twin challenges of research governance and reflexive research practice can benefit from engaging with the notion of public value.Coproduction can act to enhance the effectiveness of research by making it better informed by communities' preferences and needs, with communities then contributing to improved outcomes and achievable solutions (Ostrom, 1996).For this to be sustainable, then, we need to learn from governance practices engaged in reformulating power structures.University research governance has remained largely immune to engaging with communities in determining their research priorities, values, and impact.The foregoing discussion proposes a radical reevaluation of the nature of engagement with communities.The examples discussed demonstrate enthusiasm for harnessing the wealth of research expertise within universities to benefit communities directly.Whilst institutional research governance practices currently embed decision making within the institution, coproduction suggests a broader definition of public values which go beyond individual research practices to considering the importance of bringing communities into the institution to effect change.
Turning to reflexive research practices, the space in which to question and experiment with research practices is currently limited.The answers we may find may not be expected and most probably will be challenging.Thus, though coproduction in research is an admirable and currently resonant subject of debate, researchers should not be complacent about its use and about the significant changes it demands for research with communities in both governance and practice.We echo Robinson and Tansy, who note the 'overriding' importance of learning "how to negotiate the boundaries and terms of engagement between the research and partner communities" (2006, page 159).The outputs of coproduction move away from the measures of scholarly success that we are increasingly judged by, towards those that may have community and public value.Working to produce community-focused inputs requires time and has to be acknowledged alongside other more traditional forms of output.
Coproduced research inherently reconceptualises the role of the researcher in working with communities.As a methodology, it may not be the focus of each research project.Furthermore, we acknowledge that the current system of determining research value is unlikely to alter fundamentally.However, we have highlighted tensions created between current understandings of community participants in research and coproductive approaches.We suggest a more substantive engagement with coproduction which offers an ethical stance for researchers interacting with communities that challenge bureaucratically limited notions.It is not a singular method to be applied and will vary according to the research issue or the community engaged with.We would be sceptical of applying this framework to encourage new and more far-reaching forms of university bureaucracy but would echo Dyer and Demeritt (2009) in suggesting that research with ethics approval is not the same as ethically grounded research.
Coproduction offers not just a model or a set of insights for individual researchers seeking to extend research in collaboration with communities but contributes to an ethos which underpins the role of universities in creating public value.We have highlighted key debates of reengaging communities away from the public harm model, community presence in research governance, and allowing greater space for engagement across the researcher-community divide.Thinking more closely about public value points to a different model of creating research impact based on coproduction.This is a challenging and long-term project.This paper has provided a set of considerations in the ongoing debate about coproduction as an agenda for research that offers a differing perspective on the public value and importance of research with and for communities.More importantly, we challenge the ways that communities have become framed in research governance ethics, arguing that public value debates must value communities as partners rather than narrowly frame their involvement through ideas of harm minimisation.There is no one answer to this, but the space to explore this is not provided through the current regulation of research ethics.Reevaluating the membership of research ethics committees may be a starting point.There is no singular answer, as "what works best will vary according to context and there is a need to continue to experiment with and to monitor effectiveness of different forms of coproduction" (Martin, 2010, page 217).There is, however, space to explore this through reevaluating prevailing attitudes towards publically valuable and valued research.