Book Review: Knowledge, Policy and Power in International Development: A Practical Guide

This review appears in the Evidence and Policy journal, where it is now available free online (after I protested about the scandalous, rip-off $30 they were charging). Or you can just read it here. Note to self: in future, I will not write anything for journals that are not open access (thanks to Owen Barder for that suggestion).

In recent years, the public and policy debate over climate change, ‘climategate’, and the debacle of the Copenhagen Summit (and seemingly the wider UN negotiations) has brought home the tenuousness of the links between knowledge and public policy-making. ‘Do the research and they will come’ is clearly not a credible doctrine. Knowledge, Policy and Power, written by a group of researchers from the Overseas Development Institute, tackles some important aspects of these links, building on ODI’s strong track record on the interface between research and policy-making.

The book has good instincts – sceptical of all things linear, of researchers claiming to know more than they do, stressing the importance of values, beliefs, assumptions, taboos and other group pressures, hidden power  and in/exclusion in what are often portrayed as neutral processes of research and debate. There is ample discussion of the relative strengths and weakneses of different kinds of knowledge, whether derived from practice, ‘pure’ research or the people themselves.

Knowledge, Policy and Power argues that four key dimensions need attention in understanding how research translates (or doesn’t) into policy:  the political economy of the knowledge-policy interface, the actors who engage at it, the types of knowledge used and the role of knowledge intermediaries. It devotes a chapter to each of these, and concludes by summarizing its ‘core messages’ as:

1. Systematic mapping of the political context is necessary to improve the success of knowledge-policy interactions. Adopting the position that ‘it’s all down to political will’ is not only inaccurate but also counterproductive.
2. Understanding the role and behaviour of actors is not a simple matter of imputing self interest, but of considering the interplay of actor interests, values/beliefs and credibility and the power relations that underpin these.
3. Research needs to be complemented by other forms of knowledge, based on local conditions and practical experience.
4. Anyone working in this field as a ‘knowledge intermediary’ needs to think through a range of possible approaches to ensure their role is effective.

Along the way, it scans a vast literature to cull numerous useful typologies – of states, schools of thought, influencing factors, forms of knowledge etc, which can provide useful tools for those seeking guidance. The chapter on ‘facilitating knowledge interaction’ is the most practical and useful, setting out and discussing a spectrum of roles for ‘knowledge intermediaries’ (which I guess includes people like me), from low level ‘informing’ to ‘engaging’ to ‘building adaptive capacity’. With sensible guidelines on how do decide which approach to use in a given situation, it came closest to fulfilling the ‘how to’ promises of the book.

For the book claims to be a practical guide, which brings me to my first of three main criticisms. It isn’t very practical. The style doesn’t help: Firstly the language is variable, (chapters feel like they have been written by very different authors), but veers overall towards the opaque, with the verbiage of post-modernism (plural contexts mediated by contested discourses etc etc) scattered liberally over the text, seriously blunting its ability to communicate a clear message. Even the (very welcome) case studies seem too abstract! Example: ‘The difference between an informal designation process (Viet Nam) and a dual system where birthright and elected leaders share control (Morocco) is manifest in the degree of regulation and openness.’ Yeah, right.

That may be why, although I had regular glimmers of recognition and the odd wry smile, I had no ‘aha moments’ when reading this book. That is unfortunate– I think revelatory ideas are probably in there somewhere, but are so buried beneath the dense language, that several readings would be required to uncover the gems, and few people will have sufficient time or patience.

My second problem with Knowledge, Policy and Power is the alarming extent to which it blurs (or more accurately, ignores) the boundaries between research and advocacy. The book recommends that researchers consider ‘shaming techniques directed at veto players’ and ‘building wider movements’ as part of their day job.  It contrasts the Brookings Institution (high credibility, but limited advocacy role) with the much more overtly partisan and activist Heritage Foundation, and the authors seem to prefer the Heritage model, but don’t discuss the costs of doing so. The section on credibility is rather weak and ignores the issue of reputational damage.

I work for an organization which specialises in ‘research for advocacy’, but even I was alarmed by this – if research organizations veer too blatantly towards activism or ‘policy-based evidence making’ they risk reputational damage that can be close to permanent. Instead, I would have liked to see much more discussion on the kinds of alliances researchers can make to improve impact, while preserving their reputation, and the challenges they face in forming those alliances (for example NGOs typically work on much shorter timescales than researchers, resulting in much mutual frustration). The book seems to assume that researchers can do it all – they can’t, and nor should they.

My final point is that Knowledge, Policy and Power includes only passing reference to shocks, or ‘critical junctures’ as they are categorised in Why Nations Fail. The discussion portrays a largely steady state world of research, engagement with policy makers, and civilised debate, but in advocacy terms, readiness for such junctures is all. Very often, it is scandal, failure, crisis and disaster that drive change in policy, and that carries important implications for researchers and advocates. The most obvious is that when a shock hits, researchers should be repackaging existing research to show its relevance to the current crisis and making every effort to get it into the hands of policy makers, even if that means temporarily abandoning the cherished five year research programme. A discussion on the use of research before and after elections would have provided another excellent example of influencing in practice.

Overall, I think there is enough in here to warrant close study by researchers seeking to improve the policy impact of their work, but be warned – you will have to work at making this book produce practical guidance.

How should our influencing strategy vary with the kind of state we’re working in?

Despite the deeply unimpressive response to my last attempt (on top killer facts - not too late to chip in), I’m willing to give you another chance to provide us with unpaid consultancy crowdsource some useful ideas. This time it is helping us think through how an INGO’s influencing strategy at national level (whether through advocacy, programming or both combined) needs to adapt to the institutional environment and in particular, the nature of the state. To do this, we borrowed a handy 2×2 matrix from our humanitarian colleagues, categorizing states along two axes –Willing-Unwilling and Able-Unable (see pic).

‘Able’ refers to a state having the resources and governance structures to be effective (in delivering sustainable development).

A ‘willing state’ is one where a significant part of the state apparatus wants to deliver sustainable development and is willing to engage and involve active citizens.

Yes, I know these are very crude categories: states and countries are not homogeneous, and different strategies suit different issues, sectors and target institutions. But bear with me – the idea is to help us understand the different political contexts in which we work, and how we need to be organized for maximum impact.

So let’s unpack the four quadrants. For each one, I’m listing some tentative candidate countries, some general characteristics of the most suitable approach to influencing, and some specific strategies that might suit the political context.

1. Able & Willing
Possible candidates: Brazil; South Africa; Mexico; India? 

General Characteristics: In Able/Willing states, you can employ the full repertoire of influencing strategies, supporting civil society organizations to make maximum use of the ‘invited spaces’ offered by the state, but also supporting more confrontational approaches to create new spaces where necessary. But what works best?

Possible Influencing Strategies:
Support civil society strengthening and activism
Engage with countries’ role in the world
Engage publics (incl. using traditional & digital media) – middle class & poor
Convening and brokering discussions between different sectors (state, civil society, private sector, media, academics, faith-based etc)
Strong evidence base & research
Private sector – engage positive actors & push for regulation by state
Use legal system and test cases

2. Able & Unwilling
Possible candidates: Russia; China; Indonesia?

General Characteristics: People-on-the-streets style activism is likely to be counter-productive, but often the state technocracy is amenable to arguments based on evidence, especially when conducted through respected (state-approved) institutions. Beyond that, what else is possible, especially to strengthen citizens’ voice?

Possible Influencing Strategies:
High quality evidence & research
Partnerships with respected think tanks
Programmes that demonstrate best practice
Influence private sector CSR and encourage investment best practice
Support civil society space

3. Unable & Willing
Possible candidates: Haiti; Zambia; Ghana; Bangladesh; Kenya; Nepal; Mozambique; Nigeria?

General Characteristics: What do you do when the state’s door is open, but there is nothing much behind it? It’s all very well to support demands for change, but INGOs may also have to build the supply side – working with the state at local or national level to enable it to respond to those demands. Plus what’s the right way to engage with non-state actors to build state capacity in the long term (rather than undermine it by creating parallel systems)?

Possible Influencing Strategies:
Build civil society capacity & active citizenship
Programmes that could be taken to scale
Influence donors
Brokering role with private sector
Support communities in defending against abuses
Technical/advisory support to local/national state
Engagement with important non-state actors (faith-based, traditional authorities, other)

4. Unable & Unwilling
Possible candidates: DRC; Afghanistan; Zimbabwe; Ethiopia; Pakistan; South Sudan; Mali; Somalia; Yemen; Egypt

General Characteristics: The most difficult environments in which to do influencing (or pretty much anything else, apart from selling arms). We can support basic ‘bearing witness’ style work, and engage with non-state actors such as aid donors, but what else is possible to build a brighter future in some pretty dark places?

Possible Influencing Strategies:
Donor engagement
Humanitarian advocacy
Bear witness
Help provide support and “cover” for civil society?
Engagement with important non-state actors (faith-based, traditional authorities, other)
Concentrate on building next generation (eg work with student leaders)

So does this resonate with your reading/experience of influencing in different polities? Or is it too crude and generalised to be useful? Over to you…….

How can you do influencing work in one party states?

Fascinating conversation during a recent visit to East Africa about how we work in Rwanda and Ethiopia, both arguably effective but authoritarian states, with little time for ‘people on the streets’- style campaigning. Does that mean it is impossible to influence the state’s policies and practices? Definitely not, but you need to do things differently, in terms of how you think, your model of change etc. The discussion echoes similar ones with our staff in Russia and Vietnam. So here are some off-top-of-head thoughts about NGOs working in (for want of a better word) one-party systems.

1. One party states can be more open to robust evidence than multi-party ones. Disturbing, I know, but these governments often appear more interested in hard evidence than their more democratic counterparts. Petitions and emails won’t do it – you need data, though individual impact stories can work quite well too. Let’s not get too starry-eyed though, there will be areas that are off limits, and you need to know what they are.

2. Research will have more influence if it is done by an organization the government trusts – such as the parastatal Academies of Social Science in Vietnam and China and major Moscow-based universities. In any system, the messenger matters as well as the message, but these governments seem to specifically use such parastatals as channels to gather new knowledge.  Such research institutions are also often hungry for connections with external agencies so you’re pushing at an open door.

3. Your political economy analysis had better be good: understanding how decisions are taken in Addis or Moscow is absolutely critical. Personalities, relationships (both personal and professional), the histories, cultures and incentive systems of different leaders or ministries. Until you have a confident grasp of these, your influencing is highly likely to go wrong (always with the risk of backlash).

4. The importance of symbolism, pride and sovereignty. Traditional NGO name-and-shame tactics are as likely to lead to a counter-productive backlash as a breakthrough.  Instead, offer things that work. Find the Achilles’ heel that the government recognizes as a weak spot and focus on that. They can be surprisingly grateful for suggestions.

5. Alliances with non-state actors: in terms of citizens, most authoritarian regimes are keenly aware of issues of legitimacy, if not of direct accountability, and acts of citizenship are far from absent even when elections are either not held, or perfunctory. Building alliances with other actors (academics, churches, private sector) is just as important as in more open state systems.

6. Online campaigning and coalition work often offers more space for activism than marching on the streets. Cultivate an avatar with a particular niche, or enjoy safety in numbers working as part of a group that is harder to divide and conquer online. The use of anonymous websites, facebook groups and twitter accounts for example can be effective in creating public concern beyond the immediate community being affected. These tools can help raise awareness and motivation for people to act on the ground.

7. Persistence matters. You need a slow, steady, drip drip drip of information, invitations to roundtables, offers of secondments for mid-level officials who are the doers and internal influencers of their bosses, but not the deciders. 

8. Play to your advantages – use an international front person when talking with international organisations re the policies of the state in question. Go local when dealing with ministries – you will get further if you avoid antagonising national sensitivities.

Feel free to add your own thoughts

How to turn knowledge into policy (without losing your job)

Together with Martin Walsh, our team’s research methods adviser, I’ve been browsing through some of the literature on how to ensure our work has impact……

After a year in which Britain’s top drugs adviser, Professor David Nutt, was sacked by the Home Secretary (interior minister) for overstepping the line between providing advice and advocating specific changes to policy, you’d be excused for thinking that in the UK at least, “evidence-based policy-making” was more rhetoric than reality.

Prof Nutt is an academic, and academics in the U.K. are paid by government to produce the evidence that policymakers need, not to contradict their favoured policies. Development practitioners, thank goodness, have a little more leeway. NGOs and think-tanks use research to lobby governments and international agencies, and 2009 was a bumper year for them telling one another how to do so more effectively. Fred Carden’s ‘Knowledge to Policy: Making the Most of Development Research’ is the fruit of a study by IDRC in Canada (kudos to them for making it downloadable) that was originally designed to evaluate and enhance the influence of its own research on public policy in developing countries. Eight years, 23 case studies and several workshops later, IDRC has not only improved its own practice but also shared its hard-earned knowledge with the rest of us. The first section of the book summarises the findings of the study and is crammed with analytic typologies and tips on how to catch the ear of even the most resistant of policymakers. The second section comprises potted versions of the case studies, and the third and final section describes and reflects on the methodology of the evaluation itself.

Fred Carden sensibly concludes that the best way to start is for a would-be researcher to identify which of five ‘policy contexts’ best describes her/his situation, and then adapt research methods accordingly. They are:

1. Clear Government Demand
Implications: Give thanks. A relationship of trust between researcher and policymaker is critical. Researchers need to anticipate issues so their advice is ready on short notice. This requires researchers to be working on issues before they trigger policy interest and to think through policy implications of their findings before getting the policymakers’ call for help. It also requires clarity in communicating evidence and a reputation for delivering reliable work.

2. Government interest in research, but leadership absent
Implications: Think about issues beyond the research itself. What are the institutional and organizational implications of the evidence: if this evidence is going to be used, how will it be used? Where will decisions need to be made? What policy and regulatory changes might be implied and what effect can these have in other quarters (i.e. who are your likely allies and enemies)? Focus on your communication with decision makers, but you may also need strategies for ensuring that the interested public, that is people most likely to be affected by the use of the evidence, are fully engaged and able to assert influence on decision makers.

3. Government interest in research, but with a capacity shortfall
Implications: Propose institutional structures that draw few resources, or identify an economic pay-off that meets a priority need of government, or secure other sources of funds for implementation. You need to exert special communication skills to turn the subject into a high priority issue, perhaps by mobilizing public opinion behind their efforts.

4. A new or emerging issue activates research, but leaves policy makers uninterested
Implications: Focus on other communities that can help to promote the evidence and its merits to decision makers, e.g. advocacy groups, the media, affected communities, the private sector, and educational institutions. Stress the economic and social rewards of research implementation..

5. Government treats research with disinterest (sic), or hostility
Implications: Prepare for the long haul. You will need a strong sense of purpose and commitment to the future. Your research is preparation for a potential change of political interest. It may even be counterproductive to advocate too strongly for a course of action fiercely opposed in policy circles, and better to outlast the resistance and stand ready to seize an opening for influence when it appears.

Judge for yourselves whether other studies and guides provide better advice:

- a recent paper by Andy Sumner and colleagues at IDS, ‘Making science of influencing: assessing the impact of development research’

- the many good things that have come out of the Research and Policy in Development (RAPID) programme at ODI. One of the latest of its products is a briefing paper by John Young and Enrique Mendizabal on ‘Helping researchers become policy entrepreneurs’, summarizing more than five years’ work (involving more than 50 case studies) on ‘understanding how policy processes operate in the real world’, culminating in the development of the RAPID Outcome Mapping Approach (ROMA) as a meta-tool (that incorporates many other existing tools) for translating research into policy.

That enough?

Who do governments listen to? Some intel from the Oxfam GB media team

Oxfam GB’s media team is a class act, and has just done some useful research on ‘influencing the influentials’, interviewing senior figures in Whitehall, journalists and other ‘influentials’ (wonder what qualifies them for that?). Here are some of the headlines: Read More …