INTRODUCTION:

This edited collection of peer-reviewed papers explores critical challenges faced by organisations and individuals involved in evidence-informed development through a diverse set of case studies and think-pieces. In this chapter we briefly set out the foundations of the trend in evidence-informed decision-making and reflect on a fast-changing development knowledge landscape.

The dominant themes emerging from the contributions provide the structure for this chapter, including: building networks and partnerships; contextualisation of knowledge and power dynamics; and modes of knowledge brokerage. An analysis of these themes, and the respective roles of researchers, non-governmental organisations, large programmes and policy actors, suggests that a common thread running throughout is the importance of social relationships.

We find that the social and interactive realities of mobilising knowledge comprise several layers: (i) individual and collective capacities, (ii) individual relationships, (iii) networks and group dynamics, and (iv) cultural norms and politics, which are all key to understanding how to make evidence really matter.

Evidence-informed decision-making in a post-truth world

Given the current concerns around post-truth politics and fake news it is worth reminding ourselves that the trend for evidence-informed development expanded considerably over the past three decades. It emerged in the 1990s in health as an outgrowth of evidence-based medical practice and can now be found in virtually every area of development policy and practice. It has been the subject of a plethora of books, reports, case studies, journals, campaigns, networks, organisations, training programmes, frameworks, principles and methods. There are many different related disciplines, to evidence into policy, knowledge mobilisation, research uptake, impact evaluation, knowledge management and organisational learning. Despite all of this work, however, progress in how well evidence informs development policy and practice is at best uneven, and some commentators are suggesting it may become increasingly challenging (Economist 2016). As initiatives such as the RAPID (Research and Policy in Development) programme[1] have repeatedly found, evidence is necessary but seldom sufficient – and the most important factor in progressive change is political context. This is a familiar story, and reinforces the point made by the likes of Carol Weiss (1979) that the use of research in the sphere of public policy is an extraordinarily complex phenomenon and is only one part of a complicated process that also uses experience, political insight, pressure, social technologies and judgement. In international development, as with other spheres of public policy, decisions are likely to be pragmatic and shaped by their political and institutional circumstances rather than rational and determined exclusively by research. While it has been easy to share significant successes through impact awards and case studies, it has proved much harder to institutionalise any learning from these. Put simply, the development sector has continued to struggle to repeat the trick of turning research into action.

This challenge is playing out in a rapidly shifting development landscape, which has implications for how knowledge is thought about, accessed and used. Flows of information are becoming increasingly fragmented and unpredictable, with a larger and more diverse group of actors influencing policy and public debates. Digital technologies are fundamentally altering access to a vast reservoir of evidence and data, making the challenge less one of collecting evidence than one of selecting it. Researchers, practitioners, donors and policy actors (and their institutions) are competing with many different kinds of stakeholders, often with conflicting interests and agendas. Paradoxically, this growing complexity is placing ever-greater pressure on scientists and development agencies to ‘have the answer’ and to respond more effectively to policy agendas in ways that demonstrate their tangible impact. Donors too are a key driver of this discourse as demonstrated by the Economic and Social Research Council’s (ESRC) head of international development, Craig Bardsley, in Chapter 12 of this collection. There is clearly a great need and energy spanning academics and practitioners on the lessons to be learned from turning evidence into practice within such a complex setting. 

The seed for this publication was planted at the Research Uptake Symposium and Training Exchange – ResUp MeetUp – funded by the UK Department for International Development (DFID) that took place in Nairobi in February 2015.[2]  It was there that practitioners came together to break down and explore some of the challenges and successes that they have grappled with over the last decade. Later that year, the ESRC- and DFID-funded Impact Initiative for international development research[3] was launched to support grantees from the ten-year-old Joint Fund for Poverty Alleviation and the newer Raising Learning Outcomes in Education Systems programme, to collaborate and share knowledge to achieve impact. During a conference in South Africa where learning was shared between ESRC DFID research grant-holders it quickly became apparent that there were valuable stories to be shared and which needed to be better understood by a wide constituency of research to policy actors (Impact Initiative 2016). This collection, which has been co-funded by the ESRC DFID Partnership and IDS, is not an attempt to provide generalisable tools. Such efforts tend to focus on the process of evidence-informed policy as a largely technical and technocratic issue. Instead, we hope that it can in some small way contribute to new frameworks for understanding and navigating these complex spaces.

The reader will note the sheer diversity of viewpoints in this volume, from the knowledge management methodology that has been pioneered and evolved by Practical Action, to the programme-level support for maximising research impact of the Evidence and Policy Group to the research uptake approaches of ESRC DFID Joint Fund grant-holders. It reflects the multidisciplinarity of the editorial team who are themselves a mixture of Southern and Northern located practitioners and academics. Despite this breadth of viewpoints, or perhaps because of it, some common themes do emerge. These are all in one way or another related to the roles and capacities of knowledge brokers[4] as social actors. Far from being restricted to the perspective of academic producers or co-producers of knowledge, these case studies and think-pieces provide a broader analysis of what makes an effective broker in complex knowledge ecosystems. Concepts around the diverse roles of academics are already well served by the literature. Researchers acting as mediators between different groups, advocates and catalysts for social change has been observed particularly in programmes focused on citizenship and accountability (Benequista and Wheeler 2012). However, the case studies in this collection remind us that scholars do not have the monopoly on the generation and sharing of development knowledge. The key areas that emerge relate to networks and partnerships, contextualisation of knowledge, and modes of brokerage. In this chapter we explore each in turn.

What becomes clear as we do so is that there is a deeper set of layers to the social realities of knowledge for development. These social factors are: (i) The capacity of individuals and organisations in terms of knowledge and skills to engage in policy processes; (ii) Individual relationships that facilitate influence and knowledge brokerage; (iii) Networked relationships and group dynamics that connect up the supply of knowledge with the demand for it; and (iv) Social and political context, culture and norms. 

Research partnership and networks

Understanding the connections, or lack of them, between research knowledge and policy and practice is key to understanding how to make evidence matter. The links are deeply opaque at the best of times and rooted in power (Lukes 1974) – visible, hidden and invisible (Gaventa 2006). Gita Sen et al., in their example of the Fostering Knowledge Implementation Links Project (FKILP) describes the central importance of the ‘policy continuum’ in bringing together health researchers with mid-level health programme managers in Karnataka, India (Chapter 3). The success of this project resided on a deliberate attempt to ‘move beyond a unidirectional approach to knowledge transfer and uptake’. FKILP attempted to break down what they termed ‘impermeable barriers’ between researchers and policymakers. These barriers were largely overcome by creating new networks that included in their membership key individuals. 

Meanwhile, in their analysis of ESRC DFID funded research in Kenya and South Africa on tackling gender inequalities in education and poverty reduction strategies, Amy North et al., identify a neglected ‘middle space’ (Chapter 4). Low-ranking bureaucrats, school governing bodies and non-governmental organisation (NGO) workers were a neglected group between beneficiaries and national government when it came to the implementation of the Millennium Development Goals. The process for addressing this research-to-policy gap involved the well-established use of action research (Oswald 2016), which facilitated the co-production of knowledge with key stakeholder groups. The research team increasingly distinguished between ‘impact with’ whereby two-way dialogues and engaged scholarship directly led to changes in understanding and practice and ‘impact for/on’ associated with the more supply-driven passive and unilateral forms of knowledge translation which they observed can sometimes indirectly shift attitudes. The key difference in both Sen’s and North’s networked approach to engaged research to a more top-down research uptake strategy is above all else a focus on building key individual relationships. This individualised approach is not immediately obvious if one looks only at the big formal networks themselves. It represents a distinct layer of the social life of knowledge exchange, a concept we will return to later. 

Co-production of research is also a major area of learning for Pamela Juma et al., in their case study on the design and implementation of community health strategy in Kenya (Chapter 6). Locally generated evidence was successfully contextualised and incorporated into the resulting strategy primarily due to engagement of key decision-makers and managers as co-investigators in the study from the very beginning. Similar to the previous examples, this is a personal affair driven by individual champions. However, perhaps because of this, one of the biggest challenges faced was the relative lack of knowledge and skills of programme-level decision-makers to undertake research-to-policy activities such as synthesising knowledge and contextualising evidence. These intermediaries had a crucial role to play, but in many cases the human resources required were absent. This links to another layer of social reality – the importance of individual capacity, both in terms of research methods training as well as knowledge brokerage capability. 

When Sen et al. write about weak channels of communication, poor research communications skills are not the primary concern. It has been widely understood for some time now that research uptake is a process and not an event that hinges on the dissemination of a product (Lomas 1997). It is weaknesses in trust, relationships and networks that are emphasised by Sen as the key barriers to success. This somewhat contradicts much of the impact guidance from donors (DFID 2016), which seems to place the greatest importance on research communications. Meanwhile, Rhona Mijumbi-Deve et al. identified trust levels, perceived credibility and the ability of policymakers to engage meaningfully as key concerns when establishing a knowledge translation platform for community health policy in Uganda (Chapter 10). In a very different context Toby Milner sets out lessons from Practical Action’s knowledge management work and looks at deeper personal and organisational cultural and capacity issues that relate to NGO-based activists struggling to adapt to the slower more reflective process of knowledge exchange and learning (Chapter 7). Again the key issue here seems to be individual capacity to behave as an effective actor in this largely social process.

The contextualisation of research and knowledge

The process of contextualisation of knowledge is well served by the literature on this subject (Nutley, Walter and Davies 2007) and in a dizzying array of tool kits and guides. What is most striking about the treatment of this subject in this collection is that in almost all of the case studies contextualising knowledge hinges on the navigation of power dynamics that affect the demand for research evidence. In describing his project’s significance success in influencing national policy in Sierra Leone to support vulnerable children, ESRC DFID grant-holder Mike Wessells argues that action research methodology would have been inadequate without the pivotal role played by UNICEF (Wessells et al. Chapter 5). What he describes is a process that incorporates both a networked approach to social relations and the very individualised dimension of a key personal relationship. It was the research team’s close working relationship with one particular UNICEF staff member that enabled them to navigate the tricky domestic political territory. This is contextualisation built on personal relationships and not on generic stakeholder mapping exercises conducted in workshops. Or as Wessells puts it: ‘researchers who want to have a significant impact on policy should identify and cultivate a positive relationship with a well-positioned person who can serve as both a power broker and a trusted adviser’. 

Similarly, while navigating the challenges of being a group of ‘outsiders’ investigating inequalities in health systems, North et al. were able to conduct far more engaging interviews with individuals they had built personal relationships with. We see again and again this more individual relations layer of knowledge production and sharing intertwined with the group dynamics of the networked layer. Hence, Milner describes Practical Action’s knowledge sharing approach in Nepal, which constitutes sending social mobilisers into community resource centres to meet regularly with the same group of locals to identify specific challenges they face. Having established these personal relationships and earned their trust they then serve as intermediaries linking local citizens with local officials such as water and sanitation officers to further explore their concerns. 

However, getting the balance right between the focus on individual relationships and the more formal networked approach is not straightforward. Oxfam’s Duncan Green, author of How Change Happens (Green 2016), in his think-piece about the value of NGO and academic partnerships, warns against too much emphasis on engagement with a very small number of key individuals (Chapter 2). He points out that whole systems have to be tackled, sometimes making it necessary to build relationships with dozens of officials, advisers and gatekeepers to influence a minister to take a position on just one policy recommendation. Whether one places more emphasis on key individuals or on larger groups and networks, these are all still social interactions. 

In contrast to these more socially orientated approaches to supporting evidence-informed policy processes, another issue that emerged is the use of information and communications technologies (ICTs) to mobilise knowledge, or at least make it more accessible (Gregson, Brownlee, Playforth and Bimbe 2015). It will come as no surprise that Practical Action values highly open knowledge approaches, free sharing of data and digital knowledge curation as in the case of the Quinea project in Peru. Sen on FKILP and Mijumbi-Deve in Uganda also emphasise the value in using repositories and establishing digital infrastructure for institutional and programmatic knowledge sharing. However, the technology is also dependent on human resource capacity, and many of our cases reiterate the relative lack of investment in knowledge sharing and policy analysis skills. This is the capacity of individuals reasserting itself as a key social factor in the open sharing of knowledge, which is frequently framed as a wholly technical issue.

Modes of brokering knowledge

Given the overarching emphasis on brokerage running throughout this collection it is worth briefly setting out the manifestations of this activity. Brokering as a concept is covered in wide-ranging literature (Ward, House and Hamer 2009) and there is no space here to get too lost in the various conceptual frameworks and definitions. What stands out from our case studies are three distinguishable modes of delivery of research to policy and practice intermediary behaviour: (1) Direct brokering, (2) Indirect brokering and convening, and (3) Embedded knowledge gateways.

Direct brokering of evidence and learning is frequently referred to but framed in very different ways by different actors. For NGOs such as Practical Action it is absolutely clear that this is seen as part of an advocacy process. They are proud of their evidence-based approach to advocacy, which, it is claimed, delivers instrumental impacts on policy and practice. In contrast, academics are reticent to use such language and frequently argue that what they do is different (Datta 2012). 

Despite this tension between brokering as policy advocacy and as a more benign form of knowledge exchange, the distinction can be subject to debate (Pittore, te Lintelo, Georgalakis and Mikindo 2017). It is significant that few of our authors seem to try to persuade us that research production and mobilisation is value free. Wessells, for example, seems to demonstrate a research rigour not born out of neutrality in its analysis but out of a connectivity to political context. When it comes to an area such as child protection, or any other complex social development issue, there will often be a set of values driving those who produce and commission research (particularly when spending significant time in the field (Coffey 1999). Therefore, while Wessells is keen to reiterate that direct lobbying was not his team’s role, he does describe a process built on relationships where he was very directly involved in the brokering process. Likewise, those research networks involved in developing a community-based health-care strategy in Kenya are directly involved in a policy formulation process along with the politicians and sector decision-makers. Juma et al. even emphasise the importance of researchers exploiting policy windows in moving forward this process. These again are processes driven by social (and political) realities that include the capacity of individuals, culture and norms around the role of researchers and the values they hold and the wider networks to which they and knowledge intermediaries belong. 

When it comes to understandings around the role of knowledge intermediaries it is not just concerns around scholars as advocates that drives decisions around channels of communication. As Sen and Venis point out, the choice of the direct broker is a key strategic decision. Sen writes about the choice of the Indian Institute of Management Bangalore (IIMB) as the key knowledge broker for the academic and policy communities involved with the project. She argues that the IIMB’s academic reputation and perceived credibility supported its legitimacy. In a very different context Sarah Venis et al. describe the challenges of Médecins Sans Frontières/Doctors Without Borders around bridging the medical research and academic work supposedly feeding into their programmes with local innovation (Chapter 8). There was no obvious means of channelling or brokering new knowledge between these groups, and vital new understandings such as correct storage of insulin simply did not get translated into new practice on the ground. In the end, brokerage was institutionalised through new scientific days that brought researchers and innovators together in a safe space for mutual learning. These are the organisational cultural contexts and social norms that shape knowledge systems. 

Danielle Doughman et al. highlight another form of direct brokerage that involved providing technical assistance to decision-makers through synthesis and distilling of long technical documents into short and easy to understand formats (Chapter 11). They argue that ‘it was not enough to include the Africa constituencies in the decision-making structures of the Global Fund, because without technical support these constituencies were unable to effectively contribute to the Fund’s decisions and programmes’. The case study describes how technical support contributed to strengthening the voice of African constituencies in Global Fund decision-making processes and increased the interest of the African constituencies in using evidence in making their positions and other contributions to the global decision-making processes. 

Indirect brokerage has also been covered in a number of the cases studies. One of the most notable examples comes from the Evidence and Policy Group (EPG)[5] based at the Overseas Development Institute tasked with promoting ESRC DFID growth research (Chapter 9). Louise Shaxson sets out the different styles of brokerage her team deliver including: information intermediary, knowledge translator, knowledge broker and innovation broker (Fisher 2011). It is this last category, she argues, that has been most successful. Due to the sheer diversity of the research they support it has proven more productive to focus on facilitation and capacity building. This is about supporting new behaviours and skills and convening networks and events in order to improve connectivity and support mutual learning. She goes as far as to say that you must ‘facilitate not interpolate’. Whether one sees ‘innovation brokering’ as entirely distinct from direct brokering, the emphasis is still on relationships. In the case of EPG, this means building and maintaining relationships between the researchers and between key EPG staff and DFID and ESRC as well as between research producers, intermediaries and users. Again we need to recognise that there are two distinct layers here: one focused on formal and informal networks, and another on individual relationships. 

The final mode of brokering that is presented across a number of our chapters is the ‘embedded gateway’ (Green and Milner, this collection). Green argues that universities in particular have historically underinvested in specialist knowledge brokers or given them low status and insecure contracts and a lack of career pathways. This certainly sounds a familiar story to those of us who have been trying to nurture these capacities in research-based organisations over the past few years. However, dedicated demand-driven brokerage services do exist at both the institutional and programmatic level. This includes knowledge services such as the Governance and Social Development Resource Centre[6] and other help-desk and learning programmes designed to support development agencies’ use of evidence. Institutional-level knowledge brokerage is also explored by Mijumbi-Deve, whose primary concern is a chronic underinvestment in knowledge translation services in low- or middle-income countries. However, even these institutional-level knowledge services have a strong social element. The analysis of REACH PI clearly identifies the value of non-technical services such as building trust between key stakeholders, supporting evidence-use behaviours and building relationships that allow for the commitment to evidence-based policy to steadily grow. This is collective capacity to build and maintain social networks that support institutional-level commitments to evidence use. 

Conclusion

If there is one key message that you take away from this collection we hope it will be that research to policy processes are largely social. Technical capacities matter, of course, but not nearly as much as the social factors. The case for the primary importance of networks and partnerships and critical bodies of knowledge has been made many times (Georgalakis 2016). The concept of researchers as social actors is also well established, especially in the field of research communications (Benequista and Wheeler 2012) and more broadly (Nutley, Walter and Davies 2009). Furthermore, while individual research projects may sometimes, in the longer term and in indirect ways, impact on the lives of people in low- and middle-income countries, most medium- to long-term impacts arise from long-term bodies of knowledge mobilised by research-to-policy collaborations such as in the case of the award-winning Ebola Response Anthropology Platform.[7] The case studies and personal perspectives in this collection provide a window into the research-to-policy nexus, which suggests that the impact of evidence on development policy and practice is a social and interactive process built on personal relationships and social networks. The mobilisation of research-based knowledge hinges on multifaceted blends of individual and organisational capacity to engage, key individual relationships, group dynamics, culture, politics and social norms. 

It would require social network analysis in most cases to really understand research-to-policy processes and how things actually get done. In organisational knowledge management in the private and public sectors the significance of informal social networks has long been recognised (Allen, James and Gamlen 2007), and in development it has been well documented (Jessani, Boulay and Bennett 2014). Despite this social reality we do not organise or fund our institutions, whether university faculties, NGOs or consultancies, to nurture this social use of science. Academics often move on taking their contacts with them. INGOs flip-flop between policy and programme priorities (Green, Chapter 2) and donors struggle to fund cross-sector collaborations (Bardsley, Chapter 12). This is a huge contrast to the private sector: lobbying firms send a junior staffer to every meeting with the key client to ensure continuity; the hedge fund invests heavily in developing key relationships; and the supermarket buyer carefully establishes close personal relationships with suppliers. These examples may sound incongruous with the development sector but in the health sector at least there are examples of strategies for utilising relationships to leverage the evidence-to-policy interface. 

We hope that this collection provides a useful springboard from which to validate these concepts with existing methodologies and literature before exploring new methods for navigating complexity and the social realities of evidence. An understanding of knowledge systems as fundamentally social has profound implications for the current predominance of technical approaches to evidence-informed development. Unless we can be more cognisant of these social realities when designing and implementing programmes, we will never escape the general feeling of frustration shared by donors, researchers and practitioners that repeating the trick of turning evidence into action is so hard.