policy makers, international development, development, Overseas Development Institute, information technology, policy areas, London, knowledge, Institute of Development Studies, John Young, actors, academic researchers, Policy networks, Bridging Research, Public Policy, Planned Development, Intermediate Technology Publications, Research Report, National Center for the Dissemination of Disability Research, Public Administration, Global Knowledge Networks Panel, Challenge of Change, University of Warwick, Global Citizen Action, Indigenous Knowledge Systems, impact research, project proposals, strategic plans, communication Research, project planning, developing countries, Development agencies, epistemic communities, Research methodology, water policy, development practitioners, Environmental Policy Processes, policy shifts, power and politics, Edinburgh University Press, evidence, Research Fellow, NGO, Intermediate Technology Development Group, Policy project, International Advocacy, Box, Links Emma Crewe, International Food Policy Research Institute, Global Development Network, Department for International Development, Peoples Region United Nations Development Fund for Women World Trade Organisation Acknowledgements Emma Crewe, research question, International Development Research Council, Department for Water and Environmental Studies, Research Unit for Research Utilisation, T. Wallace, D. Lewis, Principles of Marketing, UK Department for International Development, South Asian Research Network, Journal of Modern African Studies, University of Chicago Press, Oxford University Press, International Development Research Council Lukes, Cambridge University Press, knowledge utilization, Kumarian Press, Link�ping University
BRIDGING RESEARCH & POLICY
Overseas Development Institute Bridging Research and Policy: Context, Evidence and Links Emma Crewe and John Young
Working Paper 173 Results of ODI research presented in preliminary form for discussion and critical comment
Working Paper 173 Bridging Research and Policy: Context, Evidence and Links Emma Crewe John Young September 2002 Overseas Development Institute 111 Westminster Bridge Road London SE1 7JD UK
ISBN 0 85003 606 2 © Overseas Development Institute 2002 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior written permission of the publishers. ii
2 The Problem: Power Without Knowledge
3 An Integrated Approach
3.1 The context: politics and institutions
3.2 Evidence: credibility and communication
3.3 Links: influence and legitimacy
4 Investigating the Impact of Research on Policy
4.1 Overall research question
Appendix research questions
List of boxes
Box 1 Why is research ignored?
Box 2 Pressures facing policy makers and researchers
Box 3 Gender policy neutralised in practice
Box 4 Simple research messages used to rationalise flawed policies
Box 5 The potential of information technology (IT)
Box 6 Policy networks
Box 7 Grassroots research has a global impact on policy making
Box 8 Using information to influence policy in Indonesia
DAWN DELIVERI DFID DGLS IDRC IFPRI IMF IT NCDDR NGO ODI SARN SNNPR UNIFEM WTO
Acronyms Development Alternatives with Women for a New Era Decentralised Livestock Services in the Eastern Regions of Indonesia Project Department for international development
(UK) Government of Indonesia's Directorate General of Livestock Services International Development Research Centre
(Canada) International Food Policy Research Institute International Monetary Fund Information Technology National Center for the Dissemination of Disability Research Non Governmental Organisation Overseas Development Institute South Asian Research Network Southern Nations, Nationalities and Peoples Region United Nations
Development Fund for Women World Trade Organisation
Acknowledgements Emma Crewe is a Research Fellow at University College London. Her most recent research involved exploring hierarchies, rituals and power in the British House of Lords. She has been a policy adviser to international development grant-makers, a Lecturer at Sussex University, and a social scientist working for the NGO, Intermediate Technology Development Group, as well as other development agencies. Her PhD from Edinburgh University on ideologies, policies and practices of aid was published as `Whose Development?' co-authored with E. A. Harrison. Her other publications range from articles on the social and economic aspects of energy development
to a forthcoming monograph on the House of Lords and an edited volume on rituals in European parliaments. John Young is a Research Fellow working on research-policy linkages at the Overseas Development Institute. The authors would also like to acknowledge the suggestions and help with this Working Paper from Maja de Vibe and Ingeborg Hovland. This research was undertaken in collaboration with the Bridging Research and Policy Project of the Global Development Network, Washington, with funding from the Department for International Development of the Government of the UK. The Global Development Network was launched in 1999 to support and link research and policy institutes involved in development. Its aim is to help them generate and share knowledge for development and bridge the gap between the development of ideas and their practical implementation. Address: Global Development Network, Suite 1112, 2600 Virginia Avenue, NW Washington, DC 20037 USA. Tel: (202) 338-6350/6351 Fax: (202) 338-6826/6831 Website: http://www.gdnet.org
Executive Summary Introduction Although many researchers have been driven by the aim to influence policy in the field of international development, research is often ignored, top-down, inaccurate or neglects the concerns of poor or marginalised people. A number of international development organisations have research programmes aimed at improving the links between research and policy, including the Overseas Development Institute (ODI). The ODI Bridging Research and Policy project aims to increase understanding of linkages between development research, policy and practice and develop simple tools for researchers and policy makers to promote evidence-based policy that contributes to poverty reduction, alleviation of suffering or the saving of lives. Research and policy defy neat separation but can be conceptually distinguished by their goals and methods; research produces knowledge (whether by, as examples, action-research or academic study); policy aims for continuity or change of a practice (stretching from the statements on intent to the development of policy in practice). In this paper we appraise current models of knowledge utilisation and policy making, as well as other literature that sheds light on the research/policy linkages, and offer an integrated approach to help investigate further. An integrated approach We argue that ideas percolate into development discourse and people learn about how the world is ordered through their practical experience of it. We have fused political interests, formations of actors, and discourses, taken account of the role played by wider civil society
and `street bureaucrats', and borrowed ideas from psychology and marketing, to create a three-dimensional approach consisting of context, links and evidence to assist the investigation into the impact of research on policy. We recommend an historical, contextual and comparative methodology, the aim of which would be to create a narrative of policy continuity and change and assess the impact of research on policy processes in particular areas. The similarities and differences between areas would shed light on: (a) how research has been shaped, used, ignored or reinterpreted by policy makers and (b) how researchers could more effectively contribute to evidence-based policy making. The following theoretical issues are likely to be important in the analysis: The context: politics and institutions The institutional, cultural and structural dimensions of international development are diverse, dynamic and overlapping. Policy makers and researchers cut across categories but their position of power, and the aims of the organisations they work for, can be identified. They are not only limited by macro political and economic structures but also the assumptions underlying them. For example, it is arguable that belief in the need for economic growth both rationalises and serves the global economic system. The way ideology is gendered reflects power relations
as well; the global division of labour with men dominating most policy making structures explains why men and women have different degrees of room for manoeuvre. Shifts in worldview where an explanatory model meets a crisis and is replaced by an alternative may take place at different levels of development discourse as a consequence of intellectual revolutions or campaigns of citizen action. But while researchers in academia and non governmental organisations (NGOs) often see themselves as championing radical causes, staff within bureaucracies tend to resist fundamental challenges to the status quo. v
Successful strategies for influencing policy makers and researchers have to take account of the various bureaucratic pressures limiting and enabling them, as well as those who commission or fund research: · the urge to simplify: due to resource shortages · `giantism': the bigger the budget, the greater the status · inflexible long-term project planning · fierce competition for funding: discouraging collaboration The complexity and diversity intensifies even further when researchers consider how policy is adapted, developed or distorted during practice. What influences policy practices will vary according to the priorities of street-level bureaucrats, local history, ideologies, and power relations. Evidence: credibility and communication The sources and conveyors of information may be as influential as the content; for example, people accept information more readily from those they trust. But it is our hypothesis that good quality research, local involvement, accurate messages and effective dissemination strategies are all important if aiming for more evidence-based policy-making. A lack of local involvement in drawing conclusions from research can lead to worrying impacts, such as children finding more abusive forms of work when child labour was banned in certain industries. The temptation to keep messages simple can be strong but they are then more likely to be manipulated by policy makers to gloss over the complexity of proposed solutions. There is no shortage of other ideas about why some information makes a mark. The element of surprise can usefully attract people's attention but not if the contents of the message are impossible for them to relate to their existing knowledge in any way. Assessing the effectiveness of particular communication strategies on particular audiences would appear to be an important part of any attempt to change policy. Information technologies have huge possibilities for communication within networks and coalitions but are not sufficient for building trust, require investment in people as well as technology, can be resisted, can accentuate exclusion, and can lead to the erosion of public space
. Links: influence and legitimacy The nature of the relationship between researchers and policy makers shapes how much influence they have over each other. Policy networks usually share some common values and outlooks and consciously work together to take advantage of policy `spaces' and `windows'. Policy makers themselves can be members of these networks, but if not directly involved in the influencing process, other kinds of support (such as citizen action) can become critical. What influences policy making will depend on the policy area and geopolitical level. Scientific arguments are important in some areas, lobbying by professional groups is vital in others. Certain patterns about the perception of expertise prevail across all policy areas; the marginalisation or appropriation of indigenous knowledge
is well documented. The reversals advocated in Participatory Rural Appraisal techniques and its spin-offs have challenged these power relations to some extent but have sometimes been merely another tool for `developers' to control encounters more effectively. vi
It has been suggested that international NGOs will have to be more accountable to the citizens they claim to represent, more South-driven when international development goals are in mind, and extremely well-informed, to be taken seriously. They have been advised to map out their legitimacy chains through systems of accountability (building structures that are representative of local constituencies where necessary) or relevant experience. Researchers who ensure they build legitimacy chains to their informants may also be less easy to ignore. Investigating the impact of research on policy Specific lessons about how different international development policy
areas are influenced by research in different coalitions are yet to be learned. Comparative research on different policy areas is needed to understand the differences between political and institution contexts, what is seen as credible research, successful communication strategies, the influence of key actors and the impact of relationships between researchers and policy makers. It is our hypothesis that research is more likely to contribute to evidence-based policy making that aims to reduce poverty, alleviate suffering or save lives if: · it fits within the political and institutional limits and pressures of policy makers, and resonates with their ideological assumptions, or sufficient pressure is exerted to challenge those limits · researchers and policy makers share particular kinds of networks and develop chains of legitimacy for particular policy areas · outputs are based on local involvement and credible evidence and are communicated via the most appropriate communicators, channels, style, format and timing This hypothesis might be usefully tested by piecing together an historical narrative of policy change. This would involve creating a timeline of key policy decisions and practices, along with important documents and events, and identifying key actors. The next, and more challenging step, would be to explore why those policy decisions and practices took place and assess the role of research in that process. To deal with these research challenges posed by an historical approach it suggested that: various interpretations of the past are critically reviewed by individual and groups of informants; conflicting narratives are cross-checked against documents and observations; and a range of methods, sources of information and theoretical perspectives are used in order to triangulate the findings. vii
1 1 Introduction Research could have a greater impact on international development policy than it has had to date. Policy-makers could make more constructive use of research and researchers could communicate their findings more effectively to influence policy. If more were understood about the context within which researchers, policy makers and stakeholders are working, the links between them were improved, and good quality research were disseminated more effectively, then better policy making may follow. A number of international development organisations have research programmes aimed at understanding the links between research and policy: the Global Development Network has recently started a three-year international research programme to explore research-policy linkages; the International Development Research Centre (IDRC) (Canada) is undertaking a strategic evaluation of the influence of IDRC supported research on public policy; and impact assessments by the International Food Policy Research Institute (IFPRI) have focused increasingly over the last few years on measuring the policy impact of its research programmes, and how it can be improved. In the UK evidence-based policy is the central theme of the Centre for Management and Policy Studies, established by the Cabinet Office in mid 1999. In the same year the Economic and Social Research Council established the Evidence Based Policy and Practice Initiative, a collaborative network of seven research units aiming to bring social science research
much nearer to the decision making process. In the field of international development, the id21 website was established in 1997 to give policymakers and development practitioners worldwide much faster access to research findings. The Overseas Development Institute (ODI) has been investigating research-policy linkages since 1999. This project aims to increase understanding of linkages between development research, policy and practice and develop simple tools for researchers and policy makers to promote evidence-based policy that contributes to poverty reduction, suffering alleviation or saving lives. As a first step Sutton described theoretical approaches to the policy process in political science, sociology, anthropology, international relations and management, and provided a 21-point checklist of what makes policies happen (Sutton, 1999). In this paper we hope to take this investigation further by explaining the problem, offering an historical, contextual and comparative approach for making sense of the relationship between research and policy, and making suggestions about how further research could test our hypotheses.
2 2 The Problem: Power Without Knowledge Although much research has been driven by the aim to influence policy in the field of international development the under-use of cutting-edge knowledge by those making decisions puzzles academic and applied researchers alike. It is clear that research is often ignored according to one study, of 70,000 research projects undertaken on education, only 70 had a significant influence in education policy
and practice (Molas et al, 2000) and researchers have attributed this to a range of causes (see Box 1). Box 1 Why is research ignored? · inadequate supply of, and access to, relevant information · researchers' poor comprehension of policy process and unrealistic recommendations · ineffective communication of research · ignorance or anti-intellectualism of politicians or bureaucrats · inadequate capacity among policy makers · politicisation of research, using it selectively to legitimise decisions · gaps in understanding between researchers, policy makers and public · time lag between dissemination of research and impact on policy · research is deemed unimportant, censored or controlled · some `ways of knowing' are seen as more valid than others Source: Stone (forthcoming) While some claim that research is being under-used, others argue that it has neglected the concerns of the poor and been top-down (Chambers, 1983) or that the research that does make an impact is misleading. Research that benefits the poor is so underfunded or undervalued that `90 per cent of the world's disease burden is the subject of less than 10 per cent of all international research on health' (Department for International Development, 2000). To promote evidence-policy making that responds to the demands of poor or marginalised people, the understanding of the relationship between research and particular policy areas needs deepening. The nature of the relationship between research and policy making has become a well-established topic of inquiry for social scientists. The linear model, pioneered by Lasswell in the 1950s and portraying the policy making process as a series of stages during which information is rationally considered by policy makers, has been under attack for thirty years. Caplan, for example, pointed out that the different worldviews of researchers and policy makers gave shape to a cultural gap preventing adequate use of research (Neilson, 2001). Clay and Schaffer's (1984) claim that `policy is a chaos of purposes and accidents' takes the critique of the linear model a step too far, however. The decision-making process may not be simple, but it is not accidental. We will argue in this paper, along the lines of Keeley and Scoones' (1999: 3233) suggestions, that it is structured by a complex interplay between political interests, competing discourses and the agency of multiple actors. This interplay is hard to describe. Taking account of different policy environments, institutional structures, and political arrangements renders the construction of a single explanatory model impossible, according to Stone (forthcoming). Processes of globalisation including fewer trade restrictions, the rise of transnational corporations and social movement
s, and the information technology revolution as well as shifts in institutional power away from the state in some countries, appear to have accelerated the rate of change and compound the problem of analysis. Furthermore,
3 making generalisations about how knowledge is disseminated and used involves a struggle with the diversity between policy areas. As examples, advocacy coalitions have worked in some contexts, epistemic communities in others; science dominates some areas, the law is more important to others. Consensus on how research is used by policy makers, or even whether it is used at all, eludes researchers in part because they define `research', `use' and `policy' differently. Research and policy defy neat separation. Researchers, policy makers and practitioners are not discrete groups; one individual can easily be in all three categories in different contexts or over a period of time. However, research and policy can be conceptually distinguished by their goals and methods: · Research aims to investigate, learn and produce knowledge by gathering information, contemplation, trial, and/or synthesis. In an international development context, that may involve action-research or academic study ranging, as examples, from a pilot project, to a laboratory experiment, a consultation exercise, a quantitative survey, a literature review, participantobservation or a participatory evaluation. It might be led by beneficiaries, development practitioners or academics from scientific and social science disciplines. · Policy aims for continuity or change of a practice, including plans and their evolution when put into practice (that is, the `how' as well as the `what' of decisions (Shankland, 2000)). `Public' policy is usually led by practitioners within organisations, and is what we are concerned with in this paper, rather than private policies negotiated at the level of communities or households. These broad definitions of research and policy including actionresearch and the implementation of policy, both of which are often left out achieve some clarity but also broaden the context of both domains. More processes, organisations, and actors are drawn into the arena, that is, the interface between research and policy, than might be with a narrower definition. On the other hand, we are posing a relatively focused question about the relationship between the two: how can policy makers and researchers make better use of research to contribute to more evidence-based policies that reduce poverty, alleviate suffering and save lives? We are not asking, therefore, what impact research has on policy in an undifferentiated sense. This question can lead to unhelpful common sense answers about the need to target your audience, the importance of finding like-minded people, or the benefit of simplifying your message. Rather, taking particular policy shifts within particular sectors, we suggest that understanding the differences between them, and accounting for the role played by research in particular cases, might extend our analysis of how research influences policy. This comparative approach could lead to much more specific findings about how particular policy areas work. At the same time, if particular case studies were to pose the same questions, the findings might be quite different but they would at least be more closely comparable. With this in mind, we appraise current models of knowledge utilisation and policy making, as well as other literature that sheds light on the research/policy linkages, and offer an integrated approach to help identify key research questions and investigate further.
4 3 An Integrated Approach What are the limits of existing models of policy making? While the linear model, whereby knowledge feeds into a logical sequence of decision-making phases, has few defenders anymore, other rational models continue to hold sway. Incrementalism derived from Lindblom's idea that policy makers muddle through making little reforms, rather than sudden, dramatic changes is still based on a rational assessment of self-interest (Neilson, 2001). The interactive model proposed by Grindle and Thomas (1991), takes better account of the political context by proposing that decision makers are responding to pressure exerted by interested parties and economic conditions, but still relies on the idea that individual actors calculate losses and gains. Such an approach cannot take account of the relative consistency of behaviour between people with competing interests within a particular culture. Keeley and Scoones offer one way to move beyond polarised debates about agency by synthesising different ways of looking at policy change including: · interactions between different groups with differing political interests · actor-oriented approaches: such as, policy communities and networks, interfaces, actor-network, epistemic communities, entrepreneurs/saboteurs · discourse, which is an ensemble of ideas communicated through practices via coalitions, narratives, or rhetoric They fuse all three, taking account of political interests, formations of actors, and ideas, by arguing that: `structure and agency continuous and recursively interact' (Keeley and Scoones, 1999: 28). Still more experience and ideas might, however, be usefully incorporated into such an approach. The assumption that policy is determined by elites underemphasises: (a) the role played by wider civil society, through citizen action or grassroots groups, as examples, and (b) resistance, adaption and development of policy making when put into practice. The political models of policy making which Keeley and Scoones rely on quite heavily have emerged from a focus on policy makers in Europe and the US with the consequence that the differences between and within other regions are obscured. The Global Development Network plan to make use of a two pronged analytical tool for investigating research and policy scientific determinancy and political determinancy (McNeill, D. pers. comm., 2001). However, we decided to give equal weight to a third dimension: links between actors. Kitson et al. have suggested a three dimensional approach; they argue that research utilisation should be looked at on the assumption that SI = f(E, C, F) where successful implement (SI) is a function (f) of the relation between E (evidence), C (context) and F (facilitation). However, each category is narrower than the ones proposed in our paper and, as defined by Kitson et al., more appropriate to the UK setting.1 Before we outline our three dimensional approach it might be worth mentioning how it could be used. It is clear that measuring the impact of specific research findings on policy, as if linear causality between the two prevails, would be extremely difficult. As Weiss has pointed out, research is not usually directly relevant to specific policy decisions. It may achieve influence in other important ways, namely by altering the language and perceptions of policy-makers and their advisors, but such change occurs less decisively and over time as a process of, what she calls, `enlightenment'. This works by the circulation and `percolation' of ideas and concepts, as opposed to timely, hard facts and robust theories to guide policy interventions.2 Research knowledge `creeps' into policy deliberations, and policy accretes rather than follows brisk decisions. Therefore, taking a piece of research and assessing its impact on policy would be unlikely to be successful, but identifying a 1 As quoted by Nutley et al. (2002). 2 Weiss as cited by Lindquist (2001: 1921).
5 policy change and tracing the contributing factors towards it might be more fruitful. In his recent study of why the sustainable livelihoods framework became so popular with donors, Solesbury used such a tracer study to explore the role played by research within a wider context of structures and actors (Daniels and Solesbury, forthcoming). We suggest that such an approach might begin to elucidate how structures, actors and knowledge operate within different international development policy areas. A synthesis of ideas from existing literature reviews, and the material in ODI's recent annotated bibliography on the subject (de Vibe et al., 2002), has led us to conclude that the most important dimensions when looking at the impact of research on policy in this fashion are threefold: · Context: politics and institutions. To analyse how the context shaped research and policy making it is necessary to take account of the political and institutional structures. The most important variables are likely to have been: the interests of key policy makers and researchers, structures and ideologies they were limited by, whether the policy changes were reformist or radical, how organisational pressures operated, and to what extent policies were adapted, developed or distorted when put into practice. · Evidence: credibility and communication. Investigating the impact of research findings raises questions about the credibility of the research that made an impact (how it was gathered, who by and whether it was perceived as accurate), and how it was communicated to policy makers. · Links: influence and legitimacy. When assessing how much research and researchers influenced policy, it would be important to find out the identity of key actors, the roles played (supportive or blocking), the links between them, and the extent to which the research methodology gave the findings legitimacy. The three reasonably independent clusters that make up the integrated approach are explained in greater detail below. 3.1 The context: politics and institutions Much of the literature on knowledge utilisation and policy making is concerned with the United Kingdom or United States. When researchers work within their own countries, the context can be taken for granted to some extent. However, when considering international development and, therefore, most of the nations in the world, whether they are aid recipients or donors the institutional, cultural and structural dimensions become fantastically diverse, multi-layered and overlapping. All possible kinds of institutions (state, market, civil society) operate in different modes (co-ordination, competition, co-operation respectively (Robinson et al., 1999)) and impinge on International Policy
making in different ways. Even within one institution, it is not always easy to pinpoint one set of specific policies. Robertson points out that in international politics `the art is to say as little as possible as convincingly as possible' (Robertson, 1984: 109). The UK Department for International Development's aims are expressed rhetorically in the White Paper, and where money is spent reveals a great deal, but like any organisation different parts of it follow different, and sometimes conflicting, principles. Rather than trying to describe the political and institutional context of international development, we will merely highlight some of the political and institutional structures that shape (and are in turn shaped by) policy making and research. 3.1.1 Institutional settings: policy makers and researchers Policy makers are elusive as a category. For a start, all but those in the most senior positions tend to deny that they are `policy-makers': more often than not staff are trying to influence someone higher up. Also, the policy makers in one context, Nigeria government Ministers for example, become
6 advocates in another, trying to influence the World Bank or International Monetary Fund (IMF). Donors at international (multilateral) and national (bilateral) levels are clearly making important policies but they also try to influence each other. And influential policy-makers are not only found in donor agencies, but in financial institutions
, NGOs, businesses and so on. International public bodies, civil society organisations and multinational corporations have complicated and dynamic relationships with `the public'. Furthermore, when accounting for international public opinion, making neat oppositions between attitudes within aid-receiving and aidgiving countries break down if you consider that the rural versus urban divide may have greater significance to peoples' world view, and position within power structures, than national boundaries. How the macro political and economic systems impinge on particular policy structures and processes is likely to be a part of any policy analysis. It is a challenge to keep up with the pace of change. An assumption that the state is the most significant policy making institution still holds sway in much development discourse. Ferguson argued that the `development apparatus' in Lesotho is a machine for reinforcing and expanding the exercise of bureaucratic state power.3 However, it is becoming clear that the power of state is receding in some countries while civil and corporate agencies are gaining ground. Understanding the relative power of different institutions, and how it shifts over time and in different contexts, is important when: (a) piecing together how political context shapes policy making, and (b) identifying the key policy makers. It might also be worth considering what impact advocates or researchers can have on the policy context themselves. For example, leapfrogging the nation-state, and approaching international institutions when trying to influence policy makers, can undermine democratic politics (Rai, 2002). Understanding the gulf between researchers and policy makers also requires an analysis of where researchers operate. Academics are assumed by some practitioners to be working in `ivory towers', removed from decision-making and out of touch with practical realities. But research is not simply carried out by organisations dedicated to investigation; most implementing and funding agencies, including those dominating policy making, have a research capacity or even department. Even NGOs engaged in service-delivery at the grassroots are `researching' when piloting new ideas or monitoring projects even if their consultation of theoretical work may be limited. So, identifying the researchers and how they operate can only be done on a case-by-case basis. 3.1.2 Room for manoeuvre Policy makers and researchers are not only limited by political, social and economic structures, but the assumptions that underlie them, even if they push these limits on occasions. They may take actions and use ideas that attract least criticism or the ones that they are most accustomed to and help to make sense of their reality. These structural constraints influence the context of policy making on all levels as examples: · international financial institutions make assumptions about the likely impact of debt cancellation on future lending practices · at the national level, the liberalisation of para-vet services in Kenya have been obstructed by government policies driven by the economic interests of vets · local politics jeopardised the distribution of food aid in Goma (then Zaire) when it was appropriated by some of the perpetrators of genocide in order to attempt further attacks on Rwanda 3 As cited by Crewe and Harrison (1998: 1801).
7 How power structures impinge on particular policy areas will be a matter of debate if not controversy. Understanding policy contexts involves arbitrating between a range of interpretations that will be informed by different values and interests. It is likely that the room for manoeuvre of policy makers and researchers is an uneven experience partly because it is gendered. Research undertaken by Latin American intelligentsia has been occupied with public space and modern production both associated with masculinity because men have controlled discourse (Franco, 1994). A wider global division of labour with men dominating policy making structures in many sectors explains why men and women have different degrees of room for manoeuvre, and partly accounts for the tokenistic content of much gender policy. It is widely acknowledged that gender policies often reflect flawed assumptions about women and men rather than careful research on their respective roles; for example, the idea that men are more technically minded flies in the face of women's rich knowledge of technology. It is partly women's subordinate position within communities and households that leads to the invisibility of their knowledge and work. It would be worth finding out how the links between research and policy are gendered in particular policy areas. Macro-structures and values continually change, even if some components move more slowly than others. It has been widely argued that the economic theory that had the greatest influence on development planning the Keynesian idea that planned development relies on the state to redistribute wealth has waned, and neo-classical economics has been revived. As power shifts away from the state towards the private sector
, it is inevitable that the ideologies used to justify the position of corporations have moved at the same time. Monetarism with its aims of low inflation, privatisation, welfare cuts and reducing trade union power at any cost made its way into the international development agenda. Sustainable economic growth is dependent upon an `undistorted, competitive, and well-functioning market', according to the World Bank in 1991 (Escobar, 1995). This shift reminds us that the World Bank is still dominated by economists and, therefore, an assumption that development is a mainly economic process. Researchers have produced a great deal of evidence to indicate that welfare cuts in poorer countries intensify the poverty trap, but neo-liberal policies persist. In addition, received wisdom continues to dictate that growth is an essential ingredient of poverty reduction, and both international development agencies and financial institutions advocate improving poor countries
' capacity to trade rather than negotiating trade agreements that are more favourable to poorer countries.4 Economic growth remains central, despite the evidence of social and environmental costs, because global financial systems depend on it. Accepting radical limits to growth would require a fundamental shift in assumptions about costs and benefits across the globe and between generations. 3.1.3 Reform or revolution? It might be revealing to look at policy decisions in terms of whether they are routine, incremental, fundamental or emergent, as Lindquist (2001: 1921) suggests: · routine: decision-making adapts existing programmes to new situations and entails the least scrutiny and debate · emergent: at the other end of the spectrum, goes beyond a questioning of underlying assumptions to a broad new vision 4 New ideological assumptions have led to the incorporation of strong social capital, pluralistic governance, new forms of partnership, and more legitimate civil society organisations into the development agenda. See Edwards (2001: 23).
8 We can hypothesise quite easily that staff within bureaucracies resist fundamental and emergent challenges, with their expensive and time-consuming debates and potential conflict, more vigorously than proposals for incremental change. Many researchers in academia and NGOs, on the other hand, see themselves as championing radical causes and challenging the status quo. Pross distinguishes between `sub-government' and the `attentive public' within policy communities whereby the former usually has a vested interest in the existing order while the latter tend towards criticism, creative ideas and novel approaches. Change can be inconvenient to bureaucracies and so challenges may be treated dismissively (Pross, 1986). As Chomsky points out, policies can consist of what he calls a `framework of possible thought' with various tacit doctrines that render dissent, or even debate, unthinkable. Dissident views on US foreign policy can be dismissed as anti-peace or anti-security, for example (Chomsky, 1987). But in that case why do fundamental paradigm shifts happen at all? The philosophy of science sheds light on this. Kuhn has portrayed the history of science as a series of revolutions. A scientific paradigm bumbles along quite happily making sense of the world until the puzzles it cannot solve become so numerous that it causes a crisis. A new worldview, or paradigm (most recently Einsteinian), does a better job of explaining the puzzles and so replaces (or occasionally runs alongside) the old one (Kuhn, 1970). These shifts in worldview where an explanatory model meets a crisis and is replaced by an alternative may take place at different levels of development discourse as well. When the evidence stacks up against a received wisdom an explanation for the failure of projects, for example then is it replaced with another theory? Intellectual revolutions are not the only kind of potential radicalism; pressure from below has achieved fundamental shifts in policy. The campaigns of citizen action that have met with relative success, dealing with issues like debt and landmines, have been based on a high degree of consensus within coalitions (Edwards, 2001: 6). Gaventa (2001) has gone as far as suggesting that civil society actors have challenged power relationships at all levels, influenced the World Bank, IMF, WTO and even multinational corporations, and affected a range of treaties and conventions. Other areas of campaigning trade, environmental protection and labour rights have been held back by conflicting interests. 3.1.4 Organisational pressures To understand decision-making processes that lead to policy continuity or change within bureaucracies, the pressures limiting and enabling its staff are likely to be part of the process of piecing together the context. Similarly, research aims and outputs will be influenced by pressures facing researchers within their own institutions, or professions, and by the priorities of those commissioning and/or funding the research (see Box 2). 3.1.5 Policy in practice The complexity of the context of policy making intensifies even further when the `implementation' of policy is seen as part of the process. In fact, to move beyond a linear view of project planning, it might be more useful to consider the statements of intent as well as the practices entailed in policy making. Policy statements can be adapted to local conditions, modified to take account of new circumstances, or even ignored due to conflicting interests between planners and practitioners (or different agencies) when put into practice.
9 Box 2 Pressures facing policy makers and researchers · `The urge to simplify' is overwhelming for policy makers, as Sutton (1999: 29) points out. Shortage of time and money makes it impossible to take account of diversity, nuance and contextualisation all prized by most academic researchers. Policy makers tend to devise `blueprints' with standardised ingredients for all developing countries based on simple assumptions about what causes poverty, for example rather than allowing for decentralised or contextual decision-making. In recognition of this bureaucratic pressure, lobbyists also simplify complex issues and present them as `development narratives'. These stories, such as the `tragedy of the commons' or `the woodfuel crisis', are taken up by policy makers and persist in their simple form even when evidence to the contrary calls into question their universal validity. For this reason, it is possible that research that questions received wisdom has to be presented as counter-narratives or will be ignored (for the dangers inherent in this, see section 3.2.1) (Neilson, 2001: 32). · `Giantism' is a pressure that affects almost all policy makers due to organisational rules about how power bases are established. In many agencies the bigger the budget of a project, the greater the status conferred on its manager. (Assumptions about economies of scale mean that big projects may be attractive for other reasons as well.) Individual policy makers will be partly drawn towards large-scale projects to improve their own power position within the organisation. Researchers stressing the need for smaller projects that respond to local problems and solutions may meet the barrier of giantism. · Inflexible long-term project planning is considered essential for all agencies seeking funding. Development agencies are pressed, especially by donors, for explicit, long-range strategic plans. In reality, development practitioners know that formal strategy is not a `magic bullet' because they work in volatile environment
s where conflicts between people mean that trade-offs between processes and individuals are necessary (Hailey and Smillie, 2001). Everyone involved knows that the way plans are presented (e.g., in project proposals) are fictions devised to deal with bureaucratic pressures rather than real strategies for dealing with reality. · Fierce competition for funding discourages both NGO and academic researchers from collaborative research. Different academic disciplines compete, distrust and undermine each other. Since funders and publishers often put pressure on researchers to make fresh contributions to knowledge, the emphasis of their studies is to produce original and scholarly, rather than useful, knowledge. Scholars can be fearful of making tangible recommendations. The result can be lengthy indigestible reports reflecting on complexity in the post-modern world rather than proposals for action. Research on international development is chronically underfunded and it is often the communication aims of projects that suffer cuts first during shortages. Lack of resources, and isolationism of educational institution
s in particular, can make academics aloof. NGO researchers have better links with grassroots constituencies but face even tighter budgets. Lipsky argues that policy is largely shaped by those implementing it the `street-level bureaucrats' who are often constrained by limited resources, continuous negotiation to make sure they are meeting targets, and their relationships with clients. They may be in schools or welfare departments in the US whereas in aid-receiving countries the equivalent may be a much more varied range of public and civil society organisations. They exercise considerable flexibility during implementation, often producing outcomes that may be quite different from those expected (Sutton, 1999: 8), but this may work in different ways according to the context. Conflicting priorities between partner agencies, international NGOs and their local partners, often results in gaps between what is agreed as policy, what happens in practice, and what is reported by partner staff during reviews. The result can be that the funding agencies have no idea that the supposed agreed policy has been ignored or modified by aid receiving agencies. Local conditions and power relations within communities often limit or distort the channelling of policy; it is this, rather than technical problems or shortage of information, that sometimes accounts for the failure of some projects. For example, funding allocated for women groups in Africa has been controlled by women elites, by those with good political connections, or even by men (Crewe and Harrison, 1998: 166171).
10 What influences policy practices will also be completely different in different places, between and even within countries. Keeley and Scoones (2000) show how Tigrayan versus Southern Nations, Nationalities and Peoples Region (SNNPR) regional administrative and political cultures account for quite different interpretation of discourses about participatory approaches. In Tigray there was more political space for debate and bottom-up approaches partly because of the close connections between regional actors and the ruling party. In contrast, SNNPR has been ruled by the centre since the late 1800s and has little experience of locally based forms of governance. When plans are implemented, they often look quite different. Gender equality has become an indispensable ingredient of the development agenda it is even a millennium development goal but that does not mean that equality will, therefore, be achieved when projects are implemented (see Box 3). Rather than assuming that information will solve the problem of putting policy into practice, development might be seen as arenas of negotiation for strategic groups. Box 3 Gender policy neutralised in practice Even in planning the essentially political nature of gender relations is ignored. Dealing with potential conflicts between women and men (e.g., about who controls money and new technology, makes important decisions or gets employment), are sidestepped in favour of more apparently neutral approaches that improve women's access to services. Analyses of gender, or action promoting equality, that threatens male privilege are often blocked. Since men continue to dominate most development bureaucracies, this will clearly set limits on how far policies will challenge the existing gendered status quo. Researchers tend to assume that failure to put policy into practice is attributable to lack of information. Goetz points out, however, that there is clearly a difference between `feminist knowledge', which challenges inequality between woman and men, and `data for development' which does not.5 3.2 Evidence: credibility and communication Research has been defined as broadly as possible, to include activities ranging from laboratory experiments to a monitoring consultation with stakeholders, so the purpose, method, managers, participants and outputs of the research will all vary. In general, the sources of information and who conveys it are probably as important as the content; for example, people accept information more readily from those they trust. Farmers will pay closer attention to each other's views than they will to experts (NCDDR, 1996); academics take those in their own discipline more seriously than those outside it. Those NGOs who have established their own credibility through chains of legitimacy (see section 3), a record of excellence in research, and long grassroots experience will obviously be able to use research to influence policy makers more effectively than those with a poor reputation. But attention to local involvement, accurate messages and effective dissemination strategies will all be important if aiming for more pro-poor evidence-based policy making. They will not be sufficient: we have argued that even sound policies, based on accurate research, can get distorted when put into practice; but they are a part of any sensible strategy. 3.2.1 Quality of research Research often involves trade-offs between stakeholder interests and project objectives. Since time is at a premium thoroughness may have to be compromised when ensuring that enough alternative voices are given a chance to speak. Pro-poor research implies adequate consultation with poorer people and this may have to take priority over doing an exhaustive literature review. But that does 5 As quoted by Crewe and Harrison (1998: 67).
11 not mean that low quality research is acceptable. Too often, Harper (2001) reminds us, NGO advocates have leapt from the local to global, armed only with anecdotes rather than rigorous and substantiated analysis. It is easy for policy makers to ignore researchers if they get their figures wrong. Poor quality research hurriedly or inadequately collected or analysed also means that the foundations of the message will be flawed. But it is not only inaccurate information or analysis that can be dangerous; a lack of local involvement in drawing conclusions from research can lead to worrying impacts. In a bid to ban child labour, children were withdrawn from working in some industries in Asia but this led them to find more abusive forms of work like street trading and prostitution. NGOs seem to have a more beneficial influence if they ensure local involvement and achieve high quality research by developing imaginative Research Methods
, systemising their learning processes (from local to international levels), and sharing more constructively with other agencies. Policy makers can be selective about using research. Not only was the `woodfuel crisis' a gross exaggeration, but also its rejection was based on a simplification that allowed donors to abandon all household energy programmes, despite the success some urban programmes were demonstrating in decelerating the rate of deforestation (see Box 4). Box 4 Simple research messages used to rationalise flawed policies · Policy change Multilateral/bilateral donors invest in improved stove programmes during the 1980s to save trees and cut funding during the early 1990s when they decided that the link was false. · Political context The rise in the price of oil, and emerging environmental movement, increased concern in Europe and America about deforestation and global pollution from the 1970s onwards. This was a marked contrast to the low commitment to gender and household (rather than national/global) interests in most donor organisations. Meanwhile those investing in gender programmes (e.g. United Nations Development Fund for Women (UNIFEM)) tended to focus on the productive, rather than reproductive, role of women. · `Evidence' Energy analysts made simple deductions in the 1980s, based on observation of fuel use and deforestation, that fuel-efficient stoves could save trees. To justify the decision to cut funding donors referred to: (a) observations that rural dwellers did not cut trees for fuelwood consumption publicised in a short and simple book written by a British journalist; (b) some selective and anecdotal evidence that most stove programmes had a poor record; (c) time use studies showing that when women saved time, it was not spent earning income. · `Links' Impact studies of a very small number of stove programmes, showing huge potential health and quality of life benefits at the household level, as well as studies by local researchers about links between deforestation and urban biomass, were ignored. If research is to contribute to social justice, then researchers aiming to influence policy are advised by Fine et al. (2000) to: · beware of `great stories', such as the unusual, the exotic, the bizarre, or the violent, and the search for information that merely confirm their own understandings · research `big stories' (about the historical, cultural, political, economic circumstance of one group) as well as individual/life stories (to show effects at a personal level and to bring out some variation within the big story) · draw upon a range of methods, sources of information and theoretical perspectives in order to triangulate the findings. Different research methods will reveal different versions of the story that the researcher is telling
12 · be accountable, researchers should inform the research group of how the findings will be used, invite them to critically review them, strive to stay accountable to them, and further their cause as far as possible through channels that they have privileged access
to · consider how the research might potentially be understood or misunderstood by policy-makers from different political camps It is highly likely that getting the message right requires considerable time and investment of resources. Harper advises north-based NGOs: `NGOs will have to increase their own involvement in research, build research capacity among their partners in the South, and develop more collaborative relationships with academics. In a world of highly contested and contestable evidence, NGOs must ensure that their messages carry weight and authority.' (Harper, 2001: 256). 3.2.2 Communicating research findings While there appears to be plenty of evidence that the way the message is presented matters, strategies for communication obviously do as well. We have already argued that the linear model of policy making does not work. Policy makers are not blank slates on which it is possible to write new knowledge merely by making it available. It is well established that knowledge is socially constructed, filtered through pre-conceived ideas and values. What is less obvious is how exactly information flows, is received, digested and acted upon? Stephens identifies two processes: `snowballs' (the accumulation of research impacts within policy elites) and `whispers' (the reinterpretation of research findings in broader constituencies).6 Others have described the process in terms of the limestone model (information trickles like water through porous rock), the gadfly model (information gets through because dissemination is prioritised as much as research itself), and insider model (researchers exploit links with policy-makers) (Saywell and Cotton, 1999). An understanding of how different types of knowledge make their way to policy makers in different international development areas would make communications strategies far more effective. The temptation for communicators and advocates to keep messages simple can be strong. After all it can effectively arouse public interest when trying to gain support or funds. And yet results can be disastrous when policy makers manipulate simple messages in order to ignore the complexity of proposed solutions (see Box 4). There is no shortage of ideas about other ways to help information make a mark or achieve what Gladwell (2000) calls the `stickiness factor'. He points to psychological research that shows that most people can remember up to seven-digit numbers but no more, that the presenters make a bigger impression if they outline no more than three points, and that organising more than 150 people to work effectively (e.g. as a network) is an uphill struggle. Others argue that the element of surprise can usefully attract people's attention but not if the contents of the message are impossible for them to relate to their existing knowledge in any way. It is also clear that appealing to people's moral sense can have leverage, but accusations will plainly evoke a defensive rather than constructive response. Those with little experience of, or alternative sources of information on, a subject themselves, will be most struck by visual images. As Philo (1996) argues, `seeing is believing'. Marketing relies on an analysis of what `sticks'. Its approaches became more sophisticated during the twentieth century as they shifted from broadcasting to narrow casting. This involved defining a niche market, segmentation of markets (isolating broad geographic, demographic, life-cycle, gender, income groups that make up a particular market), or micro marketing (tailoring products to specific individuals and locations) (Kotler et al., 1999). But this raises an interesting question about what 6 As cited by Berkout and Scoones (1999).
13 kind of segmentation of policy makers might be useful. Should research be communicated differently according to their level of seniority, their disciplinary background, their gender, their nationality, their age, and so on? Marketing specialists have realised that communication must be seen as a social process where attitudes and/or behaviour is changed with tangible and intangible exchanges in social relationships. traditional model
s may have targeted individuals as if they were passive, but more recent theories have pointed to the fact that identity, meaning and knowledge do not arise in the individual's mind in isolation of their environment. For example, `contemporised marketing communication theory' focuses on the complex relationships between families and households and the media and is premised on the assumption that meanings are not transferred or shared but jointly produced in social `interaction' (Varey, 2002). Social marketing has also involved theorising about what causes behavioural change. The `stages of change' model identifies ten processes that people have to move through if changing their behaviour, including consciousness raising, self-reevaluation, social liberation, and helping relationships. Since few people are ready for action-oriented programmes, time must be invested to allow for people to move through the earlier stages (Lefebvre, 2001). Does this hold true for policy makers as well; under what circumstances do they adopt values even when they are unfamiliar? Different communication channels will obviously work in different contexts. It is partly for this reason that people often play out the `official' version front-stage and give completely different versions backstage (Goffman, 1990). Research reports, newspaper articles, meetings, briefings, radio programmes, meetings, protest, the Internet, and so on, all have their place. What works for which occasion depends on many factors. Information technology has potential for communicating research findings but is not usually enough on its own (see Box 5). Box 5 The potential of information technology (IT)7 Information technologies have huge possibilities for communication within networks and coalitions with these reservations: · ITs will not build coalitions on their own because face-to-face contact has been found necessary for building trust · IT systems are social systems so have to be developed along with investment in the people who are destined to use the systems · The development and use of IT systems have been resisted by those who would lose power as information brokers · IT can accentuate exclusion; by improving access to information for some (mostly in the metropolitan core areas of the world) but leaving out others, it accentuates existing patterns of marginalisation or even creates new ones · The privatisation of information means that it is becoming a commodity not a right, that research responds to the demands of the most powerful consumers, and that public space is being eroded One piece of research has shown that using television advertising to promote social causes rarely reach high levels of effectiveness. They may encourage consumption of particular products, but this media apparently provides the wrong environment for messages that discourage consumption or other behaviours (Wolburg, 2001). On the other hand, we have already argued that `seeing is believing' and television or video has been used very effectively in other instances. For example, a multi-media venture involving development agencies and BBC World has broadcast information 7 See Volkow (1998); Peterson (1998); Castells (1993); Elliot (1995).
14 about environmentally friendly technologies to a large audience through television and radio and persuaded policy makers to take small-scale technology more seriously.8 Assessing the effectiveness of particular communication strategies on particular audiences would appear to be an important part of any attempt to change policy. It is highly likely that monitoring the effectiveness of communication strategies will improve researchers' impact in the long-term. However, this is no easy task as the time lag between research dissemination and policy change can be great. 3.3 Links: influence and legitimacy The relationship between researchers and policy makers is plainly going to shape how much influence they have over each other. Part of our hypothesis is that when they forge close personal links, with appropriate chains of legitimacy to those they represent, researchers should have more influence and policy makers could make better use of research. 3.3.1 Policy networks Policy networks, or various groupings of actors, explain how policy changes come about, according to some; it is through networks of actors that particular discourses are established or promoted (Keeley, and Scoones, 1999: 323). Networks usually share some common values and outlooks and consciously work together to take advantage of particular policy `spaces' and `windows'. To understand how networks operate, Stone, Sutton and Neilson point to different ways of mapping the actors and their networks (see Box 6). One of the key distinguishing features of different types of networks or coalitions is the position of their members in relation to the policy making process: that is, are they policy makers, `experts', street-level bureaucrats, outsider researchers and so on? A key question for research about how particular policy shifts have come about in international development is the degree to which policy makers themselves were members of the networks that had an influence. And if they were not directly involved in the influencing process, what other kinds of support were critical to the network's success? Effective policy entrepreneurs or champions will make the most of networks but will also use connections or negotiating skills, be persistent, develop ideas, proposals and expertise well in advance of policy `windows' whether brought about because of a change in government, citizen action or a swing in the national mood (Neilson, 2001: 29). Do women and men policy entrepreneurs use different strategies to gain influence and achieve different levels of success? In some instances, it is possible that the range of entrepreneurial skills cannot all be found within one individual. The spread of ideas is more effectively achieved by different types of people, according to Gladwell (2000). Three types all play a vital role in what he calls `social epidemics': · connectors are networkers, they know who to pass information to and are respected so will have influence on key players · mavens are information specialists, they acquire it and then educate others a personality type that is considered indispensable to marketing · salesmen are powerful, charismatic and, most importantly, persuasive: they are trusted, believed and listened to where others would be ignored 8 See http://www.itdg.org/html/technical_enquiries/handson.htm.
15 Perhaps researchers need to be, or use, connectors and salesmen and find `mavens' among policy makers who will pass on their message? Is there any evidence that the presence of these three different types leads to more effective use of research by policy makers? Box 6 Policy networks9 Policy communities: actors with access to privileged information inside and outside government who are highly integrated with the policy making process in specific fields, including journalists, researchers, policy analysts, elected officials, bureaucratic leaders, and those `experts' who share the same values (e.g. the Asia Pacific `policy community' in Malaysia (Camroux, 1994); or the `policy community' in Washington D.C.). Global public policy networks: alliances of government agencies, international organisations, corporations and elements of civil society that operate between and above the nation state (e.g. the Global Environment Facility, http://www.undp.org/gef/). Epistemic communities: networks of `experts' who share beliefs and engagement with a policy enterprise; professional and educational standards act as a socio-political barrier to entering these groups (e.g. Keynesian economists; or the Intergovernmental Panel on Climate Change, http://www.ipcc.ch/). Discourse coalitions: groups of actors who share a social construct and seek to impose their own discourse on the debate in different policy domains. Discourse becomes institutionalised as it dominates modes of perception and reasoning within organisations. (e.g. Israeli water policy has been driven by a farming/military discourse coalition (Jagerskog, 2002)). Advocacy coalitions: groups that share beliefs rather than interests within a sector leading to policy subsystems. Battles of ideas between different coalitions two to four in each policy community is what can cause policy change, with the dominant one usually getting its way (e.g. Jubilee 2000, or what has since become Jubileeresearch, http://www.jubileeresearch.org/). Issue networks: groups may have different values, conflicting interests, and varied solutions to a problem but they come together because they have knowledge about a particular issue (e.g. Development Alternatives with Women for a New Era (DAWN); International Forum for Rural Transport and Development, http://ifrtd.gn.apc.org/). 3.3.2 Who are seen as experts? What influences policy making will depend on the policy area and geopolitical level. Scientific arguments are important in environmental policy making, lobbying has contributed to the campaign for debt cancellation (Shankland, 2000), while economic analysis is required if working on energy. In Britain independent inquiries will usually be made up of distinguished practitioners and a token academic, in the United States a reputable Professor will often be charged with sorting out a policy problem, while researchers in Germany are often institutionalised into advisory roles and in China researchers rely on finding guanxi (social connections) when trying to influence policy makers (Stone et al., 2001: 23). Although there are differences between places and disciplines, certain patterns about the perception of expertise appear to prevail. The knowledge that makes most impact on international development policies has often been gathered by top-down methods and controlled by expatriates. Nindi asserts that many donors engage `internationally reputable consultants to ensure impartiality and lend authoritative legitimacy to their proposals'.10 It appears that knowledge is not necessarily measured and valued according to its utility for users (such as policy makers); it is predetermined by the 9 Stone et al (2001: 334); Sutton (1999); Neilson (2001); Lindquist (2001). 10 As quoted by Crewe and Harrison (1998: 89).
16 identity of the source of knowledge. Is it still the case that knowledge production and dissemination is dominated by expatriates in international development? The marginalisation of indigenous knowledge is well documented, and often rationalised by racist assumptions about the ignorance of locals and backwardness of local culture
s.11 This is apparently confirmed by the fact that `they' do not seem to `speak our language', literally and figuratively. Development organisations create their own language, rules and methods and unfamiliarity further weakens the position of those seen as `the locals'. For example, the power relationship between researchers and researcher, or local partners and donors, gives the language or more broadly, the discourse of the patron hegemony over those in the position of client. This has meant that local knowledge and innovation is often appropriated, rather than ignored, and presented as if the researcher had the idea first. The reversals advocated in Participatory Rural Appraisal techniques and its spin-offs have challenged these power relations to some extent. To give just two examples, Action Aid's Regenerated Frierian Literacy through Empowering Community Techniques Programme has enabled communities to identify and respond to problems with notable effectiveness and research with slum dwellers in Mumbai led to significant policy changes even at international levels (see Box 7). However, bottom-up approaches have sometimes merely been another tool for `developers' to control encounters more effectively. Box 7 Grassroots research has a global impact on policy making Following a survey undertaken with pavement dwellers in Mumbai, the women formed their own network of savings collectives and joined a national federation of grassroots organisations. Rather than merely submitting alternative policies to the state, they set new `precedents' to demonstrate how practice could be changed and organised mass demonstration
s of support. They designed their own low-cost housing and set up an exhibition that led to the Maharashtra government including resettling of pavement dwellers in their new strategy. Lessons Learned
about their defence of housing rights and community-based savings schemes have even been shared through regional and international networks in Asia and Africa. The Shack/Slum Dwellers International has enabled local savings groups in both Zimbabwe and Cambodia to secure land and other resources from their local governments. The focus on local processes of change, accompanied by international support from other communities as well as international agencies, accounts for their success (Patel et al., 2001). 3.3.3 Chains of legitimacy The World Bank and other donors are keen to strengthen civil society. It is assumed that `good governance' is an integral part of development and NGOs are seen as having a key role in the democratisation process, counter-balancing the power of the state. To avoid marginalisation, or being co-opted, international civil society organisations will have to be more accountable to the citizens they claim to represent, more South-driven when international development goals are in mind, and extremely well-informed. International NGOs defend their right to take positions on issues of international development as long as they were developed though `real dialogue' with southern partners. Hudson (2000) argues that NGOs would be advised to map out their legitimacy chains: when legitimacy is claimed on the basis of representation, systems of accountability need to be in place; when legitimacy is claimed on the basis of experience, the relevance of southern operational experience to northern advocacy needs to be demonstrated. There is considerable agreement on the components necessary for a successful campaign and legitimacy is usually one of them. Chapman and Fisher (1999) summarise that NGOs need: a long period of time, analysis of each level (state, market, civil society), different strategies for different target groups, collaboration, legitimacy, a narrow focus, individual champions, and grassroots 11 As examples see Hobart (1993); Warren et al., (1995).
17 support. Local involvement will have greater impact if trust is built up and it is made clear why their views are important (see Box 8). Box 8 Using information to influence policy in Indonesia12 Decentralised Livestock Services in the Eastern Regions of Indonesia Project (DELIVERI) aimed to improve livestock services to poor Indonesian farmers by reforming the Government of Indonesia's Directorate General of Livestock Services (DGLS) and its provincial and district organs. It has piloted new approaches to service delivery in four pilot districts and used the results to influence policy and regulatory changes. The project identified the critical importance of clear information reaching the right people at the right time. When improving flows of information they found that when people had faith that the research they produced would be well received, and somebody would respond to it, they would gather good quality information and articulate it well. When people felt their voices would be ignored, or they were afraid to speak, the information was often poorly considered, or chosen so as to reduce offence. Brown and Fox (2001) offer a similar recipe but add that: · power and communication gaps in civil society need bridges for people to work effectively · chains with relatively short links are stronger; that is, a few key individuals who are sufficiently close can bridge huge chasms · trust is all important to these links and the internet is not sufficient for building trust, face-to-face negotiation is often essential · being accountable to grassroots interests entails recognising, or even building, structures that are genuinely representative of local constituencies Researchers usually defend their right to represent the views of those they have studied on the basis of their scholarly `expertise'. However, it is arguable that they too have obligations to establish chains of legitimacy with their informants. While this could be asserted on a matter of principle, there may be strategic reasons for scrutinising their relationship with southern partners. It is possible that policy makers can more easily question the validity, or even dismiss, research that has been gathered with little accountability to its informants. It would be worth knowing how much the relationship between researchers and informants influences the impact of research. 12 See http://www.deliveri.org for more details.
18 4 Investigating the Impact of Research on Policy There are plenty of ideas about the impact of research on policy but they focus on public policy in the UK and the US and statements rather than policy development when putting ideas into practice. Specific lessons about how different international development policy areas are influenced by research have yet to be learned. Comparative research on specific areas is needed to understand the differences between political and institution contexts, what is seen as credible research, successful communication strategies, the influence of key actors and the impact of relationships between researchers and policy makers. From a synthesis of literature on policy making, the sociology and anthropology of development, international relations, human geography, knowledge utilisation, campaigning, media studies, management and organisational theory, social psychology and marketing, key research questions have been identified as follows: 4.1 Overall research question How can policy makers and researchers make better use of research to contribute to more evidencebased policies that reduce poverty, alleviate suffering and save lives? 1. Policy Context: politics and institutions: To what extent is the impact of research on policy making shaped by political and institutional structures, ideological assumptions and development of policy during practice? 2. Evidence: credibility and communication: To what extent did local involvement, the quality of research, and communications strategies affect the impact that research had on policy making in particular areas? 3. Links: influence and legitimacy: To what extent is research used more effectively in policy processes if researchers and policy makers share particular kind of networks and develop chains of legitimacy for particular policy areas? (More detailed questions are outlined in the Appendix). It has been argued that the power structures described in the context enable and limit certain groups, so that some may not even be aware of their own `interests' (Lukes, 1974), but individuals (such as policy makers) do make real choices. Elites may be as influenced by political and institutional factors but have greater room for manoeuvre than marginalised groups. Opportunities are created for the members of policy networks to take advantage of policy spaces or windows (described in the section on links). Actors or organisations lacking credibility will be ignored (as outlined in evidence). But policy makers can be impressed by poor quality research, if it resonates with their existing assumptions for example, and the effects of this might be detrimental to efforts to reduce poverty, alleviate suffering and save lives. Thus, it is our hypothesis that research is more likely to contribute to evidence-based policy making that aims to reduce poverty, alleviate suffering or save lives if: · it fits within the political and institutional limits and pressures of policy makers, and resonates with their ideological assumptions, or sufficient pressure is exerted to challenge those limits · researchers and policy makers share particular kinds of networks and develop chains of legitimacy for particular policy areas · outputs are based on local involvement and credible evidence and are communicated via the most appropriate communicators, channels, style, format and timing
19 This hypothesis might be usefully tested by piecing together an historical narrative of policy change in particular areas. This would involve creating a timeline of key policy decisions and practices, along with important documents and events, and identifying key actors. The next, and more challenging step, would be to explore why those policy decisions and practices took place and assess the role of research in that process. Since causal link
s cannot be assumed just because events follow each other, care has to be taken to judge what had a critical influence on whom. Ideas are often created at the same time by a number of people the zeitgeist and many will, therefore, claim credit for the good or potent ones. People may have their own reasons for remembering the past in ways that suit their image of the present (Hobsbawn and Ranger, 1983). Furthermore, peoples are notoriously selective, partial, biased and prone to leave out what they assume can be taken for granted. They are not merely recalling in a mechanical way but will be filling in gaps to make sense of an inevitability complex reality. To deal with these research challenges, the following strategies may be useful: · Finding out about the assumptions, interests, and pressures affecting each informant as well as the researchers. To test conflicting narratives and crosscheck findings, various interpretations might be tested out on individuals and groups. · Unravelling the assumptions underlying informants' interpretations and testing claims against observations, secondary sources, evidence of decisions, and other views. · Ensuring that informants know how the findings will be used and inviting them to critically review them before they are finalised. · Using a range of methods, sources of information and theoretical perspectives in order to triangulate the findings.
20 Bibliography Berkout, F. and I. Scoones (1999) `Knowing how to change, Environmental policy learning and transfer' Development Research Insights, 30: 12, June 1999. Brown, L. D. and J. Fox (2001) `Transnational Civil Society Coalitions and the World Bank: Lessons from Project and Policy Influence Campaigns' in M. Edwards and J. Gaventa (eds.) Global Citizen Action, Boulder, USA: Lynne Rienner. Camroux, D. (1994) `The Asia Pacific policy community in Malaysia', Pacific Review, 7(4): 42135. Castells, M. (1993) `The informational economy and the new international division of labour' in M. Carnoy, M. Castells, S. S. Cohen and F. H. Cardoso (eds.) The New Global Economy in the Information Age, London: Macmillan. Chambers, R. (1983) Rural Development: Putting the Last First, New York: Longman. Chapman, J. and T. Fisher (1999) The Thoughtful Activist, A Toolkit for Enhancing NGO Campaigning and Advocacy. London: New Economics Foundation. Chomsky, N. (1987) `The Manufacture of Consent' in J. Peck (ed.) The Chomsky Reader, Serpent's Tail: London. Clay, E. and B. Schaffer (eds.) (1984) Room for Manoeuvre: An Exploration of Public Policy in Agriculture and Rural Development, London: Heinemann. Crewe, E. and E. Harrison (1998) Whose Development? An Ethnography of Aid, London: Zed Press. Daniels, D. and W. Solesbury (forthcoming) Sustainable Livelihoods Approach: tracing the influence of research on policy and practice, London: Department for International Development. de Vibe M., I. Hovland and J. Young (2002) `Bridging Research and Policy: An annotated bibliography', ODI Working Paper 174. London: Overseas Development Institute. Department for International Development (2000) `Eliminating World Poverty: Making Globalisation Work for the Poor', White Paper on International Development, London: HMSO. Edwards, M. (2001) `Introduction' in M. Edwards and J. Gaventa (eds.) Global Citizen Action, Boulder, USA: Lynne Rienner. Elliot, P. (1995) `Intellectuals, the `information society' and the disappearance of the public space', in O. Boyd-Barret and C. Newbold (eds.) Approaches to Media, A Reader, London: Arnold. Escobar, A. (1995) Encountering Development. The Making and Unmaking of the Third World, New Jersey, USA: Princeton University Press. Fine, M., L. Weis, S. Weseen and L. Wong (2000) `For Whom? Qualitative Research, Representations, and Social Responsibilities' in N. Denzin, and Y. Lincoln (eds.) Handbook of Qualitative Research (2nd edition), Thousand Oaks, USA: Sage Publications
. Franco, J. (1994) `Beyond Ethnocentricism: Gender, Power and Third World Intelligentsia' in P. Williams and L. Chrisman (eds.) Colonial Discourses and Post-colonial Theory, A Reader, New York: Harvester Wheat-sheaf. Gaventa, J. (2001) `Global Citizen Action: Lessons and Challenges' in M. Edwards and J. Gaventa (eds.) Global Citizen Action, Boulder, USA: Lynne Rienner. Gladwell, M. (2000) The Tipping Point, London: Little, Brown and Company
. Goffman, E. (1990) The Presentation of Self in Everyday Life, Harmondsworth, UK: Penguin. Hailey, J. and I. Smillie (2001) Managing for Change: Leadership and Strategy in Asian NGOs, London: Earthscan. Harper, C. (2001) `Do Facts Matter?' in M. Edwards and J. Gaventa (eds.) Global Citizen Action, Boulder, USA: Lynne Rienner. Hobart, M. (1993) An Anthropological Critique of Development: The Growth of Ignorance, London: Routledge. Hobsbawn, E. and T. Ranger (1983) The Invention of Tradition, Cambridge: Cambridge University Press.
21 Hudson, A. (2000) `Making the Connection: Legitimacy Claims, and Northern NGO's International Advocacy' in D. Lewis, and T. Wallace, (eds.) New Roles and Relevance. Development NGOs and the Challenge of Change, Bloomfield, USA: Kumarian Press. Jagerskog, A. (2002) `The Sanctioned Discourse A crucial factor for understanding water policy in the Jordan River Basin', Occasional Paper 41, Sweden: Department for Water and Environmental Studies, Linkшping University. Accessed http://www2.soas.ac.uk/ Geography/WaterIssues/OccasionalPapers (1.7.02). Keeley, J. and I. Scoones (1999) `Understanding Environmental Policy Processes: A Review', IDS Working Paper 89, June. Brighton, Sussex: Institute of Development Studies. Keeley, J. and I. Scoones (2000) `Knowledge, power and politics: the environmental policy-making process in Ethiopia', Journal of Modern African Studies, 38(1): 89120. Kotler, P. and G. Armstrong (1999) Principles of Marketing, FT Prentice Hall. Kuhn, T. S. (1970) The Structure of Scientific Revolutions, Chicago: University of Chicago Press. Lefebvre, R. C. (2001) `Theories and Models in Social Marketing' in P. N. Bloom and G. T. Gundlach (eds.) Handbook of Marketing and Society, London: Sage. Lindquist, E. A. (2001) Discerning Policy Influence: Framework for a Strategic Evaluation of IDRC-Supported Research, University of Victoria, Canada: International Development Research Council Lukes, S. (1974) Power: A Radical View, London: Macmillan. Molas, J., P. Tang and S. Morrow (2000) `Assessing the non-academic impact of grant-funded socio-economic research: results from a Pilot Study
', Research Evaluation, 9(3): 172, December. National Center for the Dissemination of Disability Research (1996) Review of the literature on dissemination and knowledge utilization, USA: NCDDR. Neilson, S. (2001) Knowledge Utilization and Public Policy Processes: A Literature Review, Canada: Evaluation Unit, International Development Research Council. Nutley, S., I. Walter and H. Davies (2002) `From knowing to doing: A Framework for understanding the evidence-into-practice agenda', Discussion Paper 1. Research Unit for Research Utilisation, University of St Andrews. Linked to the ESRC Network for Evidence-based Policy and Practice Accessed: http://www.st-and.ac.uk/~cppm/KnowDo%20paper.pdf. (15.5.02) Patel, S., J. Bolnick and D. Mitlin (2001) `Squatting on the Global Highway: Community Exchanges for Urban Transformation' in M. Edwards and J. Gaventa (eds.) Global Citizen Action, Boulder, USA: Lynne Rienner. Peterson S. B. (1998) `Saints, Demons, Wizards and Systems: why information technology reforms fail or underperform in public bureaucracies in Africa', Public Administration and Development 18(1): 3760 Philo, G. (1996) `Seeing and Believing' in P. Marris and S. Thornham (eds.) Media studies, A Reader, Edinburgh: Edinburgh University Press. Pross, P. (1986) Group Politics and Public Policy, Toronto: Oxford University Press. Rai, S. (2002) `Networking Across Borders: South Asian Research Network (SARN) on Gender, Law and Governance' Paper prepared for the Global Knowledge Networks Panel, 43rd International Studies Convention: New Orleans
(mimeo). Robertson, A. F. (1984) People and the State: An Anthropology of Planned Development, Cambridge: Cambridge University Press. Robinson, D., T. Hewitt and J. Harriss (1999) `Why inter-organisational relationships matter' in D. Robinson, T. Hewitt and J. Harriss. (eds.) Managing Development, Understanding Interorganisations Relationships, London: Sage. Saywell, D. and A. Cotton (1999) Spreading the Word: Practical guidelines for research dissemination strategies, Loughborough, UK: Water, Engineering and Development Centre. Shankland, A. (2000) `Analysing Policy for Sustainable Livelihoods', Research Report 49, Brighton, Sussex: Institute of Development Studies. Stone, D. (forthcoming) Getting Research into Policy? University of Warwick (mimeo).
22 Stone, D. with S. Maxwell, and M. Keating (2001) Bridging Research and Policy, An International Workshop, Funded by the UK Department for International Development, Warwick, UK: Centre for the Study of Globalisation and Regionalisation, Warwick University. Sutton, R. (1999) `The Policy Process: An Overview', ODI Working Paper 118. London: Overseas Development Institute. Varey, R. J. (2002) Marketing Communication Principles and Practice, London and New York: Routledge. Volkow, N. (1998) `Strategic use of information technology requires knowing how to use information' in C. Avgerou (ed.) Implementation and Evaluation of Information Systems in Developing Countries. Proceedings of the Fifth International Working Conference of the International Federation for Information Processing, IFIP WG 9.4. USA: NCP. Warren, D. M., L. J. Slikkerveer and D. Brokensha (1995) The Cultural Dimension of Development: Indigenous Knowledge Systems, London: Intermediate Technology Publications. Wolburg, J. M. (2001) `Misplaced marketing: Why television is the `wrong' environment for public service advertising campaigns' Journal of Consumer Marketing 18(6): 471473.
Appendix Research Questions
Overall research question: How can policy makers and researchers make better use of research to contribute to more evidence-based policies that reduce poverty, alleviate suffering and save lives?
Policy context: politics and institutions To what extent is the impact of research on policy making shaped by political and institutional structures, ideological assumptions and development of policy during practice?
Research questions 1. How did the global, national and community-level political, social and economic structures and interests affect the room for manoeuvre of male and female decision-makers in particular policy areas? 2. Who shaped the aims and outputs of the research, how and why? 3. How did assumptions influence policy making, to what extent were decisions routine, incremental, fundamental or emergent, and who supported or resisted change? 4. How did applied and academic research influence the development of policy when being put into practice?
Information required Any political, social or economic factors which might have influenced the key decision makers. For example, in the planned ODI case studies these might include: · financial interests of international banks in relation to debt cancellation · socio-economic interests of vets in blocking the liberalisation of regulations · impact of local political interests of warring factions on food distribution Information about: · why the research was carried out · who commissioned, funded, planned and monitored it · who influenced its aims and outputs · the interests of the designers and other stakeholders Information about: · existing knowledge and values and the various actors · the old and new ideas underlying decisions · how much new policies threaten the status quo · who supported and resisted the changes and how Trace how the policies were developed, adapted or distorted as they were put into practice by, e.g., getting information about actions taken, and research used, by project partners, `street level bureaucrats', and communities
Theoretical basis Development theory; social development; international relations; gender in development (see sections 3.1.1, 3.1.2) Organisational pressures (see section 3.1.4) Philosophy of science; diffusion of innovation; social change (see section 3.1.3) Policy in practice (see section 3.1.5)
Evidence: credibility and communication To what extent did local involvement, the quality of research, and communications strategies affect the impact that research had on policy making in particular areas?
Research questions 1. How was information gathered and by whom? 2. What was perceived as credible evidence by different actors and why? 3. Did researchers segment their audience and if so, how, and did this affect impact on policy? 4. How and why was information ignored, reinterpreted and distorted and by whom? 5. Did the communicator, channel, format, style or timing of the communication play a role in influencing policy makers?
Information required List who did the research: beneficiaries, NGO practitioners, activists, academics, consultants, government researchers; was it academic/applied and before/during/after/not part of a `project' with non-research aims Information about what parts of the knowledge that was produced was valued, by whom, and how did they assess research quality
Information about whether information was tailored for particular policy makers (according to organisation, gender, discipline, policy area, etc.) and what impact it had Information about what happened to research findings within policy communities, e.g. snowballs, whispers, trickles etc. List how research findings were communicated, at what stages, and by whom, and assess what worked, what failed, when and why
Theoretical basis Research methodology (see section 3.2.1) Knowledge production, evaluation (see section 3.2.1, 3.3.2) Marketing (see section 3.2.2) Knowledge utilisation; knowledge management; diffusion of innovation (see section 3.2.2) Campaigning; media studies; communication (see section 3.2.2)
Links: influence and legitimacy To what extent is research used more effectively in policy processes if researchers and policy makers share particular kind of networks and develop chains of legitimacy for particular policy areas?
Research questions 1. What roles were played by which kind of groups and male/female individuals and what were the links between them? 2. Which women or men had significant influence over the policy? 3. How did women and men researchers and advocates establish legitimacy; did it make any difference to the policy outcomes?
Information required Information about the relationships and networks between actors. They might include: epistemic communities, global public policy networks, issue groups, advocacy coalitions; and citizen action groups etc. Identify: · who were regarded as experts in the particular policy area · who were the effective and ineffective policy entrepreneurs, connectors, mavens and salesmen · any differences in impact according to gender, nationality, race and/or class Describe chains of legitimacy between various stakeholders (e.g. whether southern partners controlled monitoring of research, or degree to which informants were consulted about policy recommendations), and assess influence of these chains on quality of policy decisions
Theoretical basis Models of policy making and change; policy networks; campaigning (see section 3.3.1) Theories on power, expertise and knowledge; marketing; socio-political hierarchies (see section 3.3.2) Role of civil society; relationship between north and south; campaigning (see section 3.3.3)
E Crewe, MJ Young