From Demand to Disinterest: Contexts for Policy Influence

Thank you for inviting me, I have been looking forward to this. I will give a brief introduction to the International Development Research Centre (IDRC) before I start. For those who do not know about us, to use the earlier analogy, I guess the IDRC works with the plankton. We are focused on building capacity for research and on the conduct of research in developing countries. Our real interest lies in building research capacity in Africa, Asia and Latin America and we are trying to increase both the volume of research and the number of researchers who are involved. Given that we are focused on research for development, one of the big questions that we have is how research influences policy, and influences development policy and technology in particular.

I will talk briefly about how we came to do our study on the influence of research in public policy and how we carried it out. I will then focus on one core aspect of the study's findings. This will bring us back to an issue already raised, which is the importance of context. We found that this was an interesting and useful way for the Centre to talk about research and the policy process. One of the reasons for doing the study was that we discovered that most of the programme staff within IDRC had a relatively unsophisticated understanding of what was meant by policy influence. The different divisions of the organisation, whether working on information and communication technologies for development (ICT4D), or environment and natural resources management (ENRM) research, or in social and economic policy, had very different views and language for talking about policy influence. This meant that the capacity within the organisation to talk about policy influence was tremendously weak, and there was a tendency to focus on the supply side with a lack of understanding of what the demand was, or of the issues which come out of a government system. In our study, we focused not so much on the demand or the supply, but on the relationship between those things: how does research relate to policy processes.

We did quite a bit of background work looking at what we had done in the past and on what was happening around IDRC, as well as at the different perceptions and understandings within the organisation, in order to find out what the starting point should be. Out of this came a framework to think about policy influence, based on the realisation that it is not so much about simply changing a policy to create a new policy, or confirming existing policies, but rather, building capacities for policy influence within the research communities is critical. In other words, we needed to find out how policy-makers understand the policy process, whether they have an understanding of how policy relates to research and whether researchers can convey what they know in ways that policy-makers can understand. Very often researchers are very poor at doing that.

We carried out our research through conducting a series of twenty-two case studies across each of the regions in which we operate and across all the domains in which we work, including environmental resources, social policy, economic policy and ICTs for development. We did most of the case studies with researchers in the countries where the work was taking place, so that we would have people involved in the research who also understood what was going on in the wider context around the case studies.

We chose case studies where someone in IDRC thought that the project had had some kind of policy influence and that the research had been useful or used in the policy process in some way. Our goal then was to understand what had actually happened. That was part of the larger goal of developing a common language within IDRC to talk about policy influence and to understand the nature of those research projects which lead to policy influence.

We developed the case studies together with the IDRC staff and our research partners. After conducting the case studies we also did preliminary analysis with these two groups. Out of that preliminary analysis, conducted through a series of workshops around the world, we came up with a series of areas where we were going to do more detailed analysis. The first area of analysis was concerned with what IDRC does, the kinds of inputs into the process, and the type of role which IDRC plays, as well as how far it mattered whether the research was explicitly intended to influence policy. The second area was about where we worked and what kind of contexts that we work in. This is the area on which I am going to focus here. The other area was focusing on the way we work in terms of networking, partnerships, the kind of research we conduct, the importance of the quality of research and issues around the timing of research.

I want to focus on context because this was an area which gave us a lot of rich material for thinking about the design of research and for discussions with researchers about whether and how they can have some kind of policy influence.

We looked at two main dimensions of context. A third dimension of context which we did not look at was where sudden political, economic or environmental changes had brought about a dramatic shift in the policy process. I do take as interesting the point which Phil made - that this is precisely when you need to have a better understanding of research and of knowledge - but we did not think that it was an area that we could learn too much about at this stage.

The first dimension which we looked at concerned the external factors which needed to be understood in order to have some influence. The first, and fairly obvious factor, was that having some level of stability in decision-making institutions is critical. Where there is a lack of stability, it is difficult to influence. Stability works at many levels. For example, the Latin American Trade Network was established so that researchers could begin to introduce some ideas, findings and knowledge with which to engage in the trade policy process in Latin America. At that point, the trade policy discussion was largely managed by international institutions and not by the countries of Latin America. One of the difficulties for the Network was the constant and rapid turnover of staff in government structures in Latin America, making it extremely difficult for researchers to work with them. They finally discovered that if they worked at the third level of the bureaucracy they would have a greater degree of success, because at that level there was stability, continuity and corporate memory on trade policy and trade issues.

The second factor, again a fairly obvious one, was whether decision-makers had the capacity to use research, seek knowledge and think critically about ideas. Whether there is decentralisation or tight control makes a difference. It does not determine whether it is possible to have influence, but it does change the nature of how we work and who we work with. Working in a decentralised and open structure in the Philippines on poverty monitoring is quite different from doing poverty monitoring work in Vietnam, where the government system is very tightly structured and where it matters which think-tank you work with as well as what you actually discovered.

The two last factors are areas of special opportunity. Whether a country is in transition makes a big difference. Being able to identify and participate in a transition in a country such as Vietnam (which was not in a political transition but was in an economic transition) and being able to provide evidence made it much easier to have an influence. This was similarly the case for countries which are under economic pressures, either from the international community or from their own citizens, to create changes in some aspect of policy. Those are a set of factors which, particularly in the design process of the research, made a big difference.

The second element of context which we found very interesting in the case studies and which has started to play a role in the design of research within IDRC, involves thinking about what kind of context exists for policy influence. This ranges from the notion of policy-maker demand and requests for information, to policy-maker disinterest. An issue which a researcher is highly interested in may hold no interest whatsoever for a policy-maker. I will give you some examples at different levels of policy windows, starting from wide open windows of opportunity where the information has been sought, and where presentation of the information is less important that the fact that it is available when policy-makers need it and in a format which they can use.

It also matters whether the research comes from people who policy-makers trust to give them the quality of research they need. This often implies situations where there have been ongoing relationships - not necessarily ones which have resulted in research being used in the past, but where information has been provided and relationships of trust have been established. One example is of a researcher who could be characterised as a research entrepreneur. He ran a research centre in Senegal and created a forum in Dakar for the government, leaders of the opposition and researchers in the economic community to talk about issues relating to economic development in Senegal. One of the few rules of the discussion was that the forum was not a place to attack or defend the particular policy positions of political parties or interest groups, but was an opportunity to exchange ideas and share information. The discussion was managed and facilitated in that way and as a result, the researcher developed good relations with decision-makers as well as the opposition, his research was trusted and he was engaged on an ongoing basis in dialogue with other researchers. He also wrote up his findings in popular papers which are sold in book stores and so on. There were a number of factors that he relied on to build a set of relationships both inside government and in other parts of the political process. These led to a situation where, when there was a demand, for example in the PRSP (poverty reduction strategy paper) process, he was asked to lead that process. His success was built on a set of relationships and on the quality of research over time which led to demand for research by policy-makers.

The next two levels concern situations where there is policy-maker interest, but either a leadership gap or a resources gap. We kept these two factors separate, even though we have only one example of a resources gap, because we think there is a qualitative difference between not being sure where leadership should be placed to make a policy change, and simply not having the human or financial resources to make that decision. In terms of a leadership gap, we found that what became important was that the researchers not only dealt with research issues and the generation of research data, but also thought about the decision process within the government system, where that decision could be housed and where those decisions could be taken. In other words, they engaged with the decision-makers and with the bureaucracy on how and where decisions would be made, as well as how the research could feed into decisions about who would take the lead in making any sort of policy change. In terms of a resources gap, the example we had was of the clean up of the Dnipro river in the Ukraine. Here the issue was clearly well understood by government; they knew that the ongoing pollution and siltation of the river was critical and was going to cause tremendous economic hardship over time. They did not have the money or the people to deal with the problem. For this reason, a project designed around building local capacities to deal with the problems of the river and identifying technologies that could be used to monitor the quality of the water in the river, became quite a critical interest and activity.

The fourth level concerns situations where there is policy-maker neutrality but a strong research interest. This is where I see most of the IDRC-supported research taking place and the level at which most researchers operate. Very often this is because researchers are ahead of the game and policy-makers have not yet come to the issue. Where this is the case, researchers are provided with the opportunity to carry out research before it is demanded, so that it is available at the point when policy-makers actually want it. The fifth and final level reflects the situation where researchers generate knowledge but the decision makers explicitly decide not to use it.

The next slides show where the projects were at the end of the study (See slide 7 and slide 8). It takes the research projects that we looked at and tries to identify where they were at the start of the project, what kind of a project they where, why they where started (demand / researcher interest) and so on. There were two main projects where there was very clear demand from policy-makers for the research. The other projects were all ICT projects, often very small projects taking place over very short timeframes, in which there was also a very strong demand. I would attribute this to the fact that ICT is a very new area compared with, for example, economic policy, education policy or health policy. As such, it was much easier for decision-makers to admit that they did not know, or were less than knowledgeable about the subject and to ask researchers to tell them than it would be in relation to the field of health or other areas of public policy where decision-makers are expected to know more.

This puts the level of policy-maker neutrality and a strong research interest as the most important level in terms of where most research sits. It is also the most risky in terms of policy influence because of relatively low decision maker interest in the issue. When you look at the projects that were at this level at the start of the project, all of them moved somewhere either up or down the chain during the life of the project. Some moved up to having a clear demand from policy-makers, including research projects on fisheries and on indigenous knowledge; others did not succeed. In the case of high-altitude mining, they did not succeed because the government did not want to introduce a change in its policy. The research was looking at whether or not mining at high-altitudes in artisanal mines in the high Andes was worse for people's health than mining anywhere else. The study found that there were very specific health problems relating to mining at high altitudes which are not found anywhere else. This was contrary to the commonly held view that people adjust to living at high altitudes so that overall nothing is more or less significant at high altitudes and the problems which could be expected were just the same as the problems we could expect at low altitude. Because of this commonly held view and because the government did not want to have to deal with the costs implied by that change of policy, the research was not used and was buried. This commonly held belief that it was ok also meant that there was no pressure from any particular sector outside of government, including from the minors, to make use of the research.

It is interesting to note that projects changed over time and that none of them stayed at the level of policy-maker neutrality and researcher interest. In these situations, where the researchers were trying to have an influence on policy, they made conscious efforts in that direction with greater or lesser success.

In terms of the conversation within IDRC about the different types of influence, one of the important things was not so much the identification of a particular context for a project, but also the nature of the factors that go along with the research. I mentioned when I was talking about the leadership gap that one of the big issues there is thinking about the institutionalisation of new policies (where they are housed, who holds responsibility; a second element is that of communication: in cases where there is direct demand from decision makers, there is limited need for sophisiticated communiqués and persuasive policy documents out of the research; where decision makers are less certain, the researchers need to ensure clear transmission of their findings in language and formats the decision makers can use; the element of advocacy is also important in some situations. Some researchers act as advocates themselves; some advocacy organizations use research as a tool to get their point across; and sometimes researchers transmit their findings not only to decision makers but also to individuals and groups that will use the findings in their advocacy efforts. What appeared important in the successful cases of policy influence was that the researchers understood the relative importance of these factors in the use of their research. So we have to be mindful of how research is presented and who it is presented to. In many cases, such as the copper mining case, we need to go far beyond the immediate policy community or decision makers, to bring the community into the debate. In this example, an NGO was concerned about the pollution of the local water source, the Ilo River, by a copper mining operation. The tailings from the mining operations were leaching into the river, polluting the water and affecting the health of the local community - many of whom depended on the mine for employment. The NGO had to find a way to address the concerns of local government and the community about employment as well as health issues. They worked at getting both the local government and the community to understand the findings of the research, so that they could then put pressure on the government to force the Southern Peru Copper Corporation to change their practices in terms of dealing with tailings from their mines. The NGO was able to take advantage of an international agreement to which Peru was party and which provided a relatively protected platform through which the bring the case forward.

In terms of who researchers direct their advocacy towards, one of the biggest challenges within IDRC is that we see ourselves as research funders and not as advocates. The next level is researchers working in countries in the South. Are they advocates or are they people simply presenting ideas that can be used? In some cases, such as the example I just mentioned, they were advocates and had a very clear agenda such as to stop the production of mine tailings and the dumping of the tailings into the Ilo river. They were able to use research as one of the tools in their arsenal to generate enough interest and pressure on the government to promote change. In other cases, the researchers are not advocates but they work with advocacy and communication groups. In a number of examples, the researchers did not try to become everything (the advocates, the communicators and so on) but they nonetheless developed their capacity and understanding of the roles of communicators and advocates as part of what they did in the research process.

What I have presented was one of the three main sets of findings within our research on the influence of IDRC-supported research on public policy. I also wanted to say a word about the use of this research within IDRC. By discussing and developing the analysis with IDRC staff and our research partners, one of the things which we discovered, or rediscovered, was that a focussed conversation on the role of research in the decision process has served to reshape the conversation within IDRC. It has given staff and our researchers a common way of talking about the policy process and the role of research in that process. It has also provided some tools to debate with researchers and other communities about whether or not a particular project has the potential to influence research. It has focused thinking on the topic, and it has also influenced practice. Most of the programmes which are intended to have policy influence (and it should be said that not all of IDRC's work is about policy influence) have now integrated a consideration of issues relating to context and the inputs and mechanisms which get used in the research process, as part of how they think about the dimensions of policy influence. It has also provided a number of stories which people can use as case material to understand what has happened in a research and policy process. Overall, the context issue is one that we have found particularly relevant and useful for talking about research and policy.

Thank you.

Comments and Questions from the floor included:

• It was noted that although none of the cases examined showed a drop of demand or policymaker interest, it is conceivable that the demand for research could fall over the projects' lifetime.
  
• It was asked how IDRC maps the policy process. A number of different elements are involved in the mapping process including the values of the evaluator, data from the country and IDRC local staff.
  
• It was suggested that within IDRC there is an ongoing debate about the role they and the research they support should have in shaping policy processes. Fred Carden answered that there is a belief among its staff that not everything IDRC does should be about influencing policy. There has, however been a shift within the IDRC more recently. Perhaps this reflects the increasing need to illustrate to those who fund the IDRC that their funding has had an impact by showing them where the research has been used. At least, IDRC is working to understand the context and the value of the research they support.
  
• It was explained that the research in this case was done ex-post, but that it would have been useful to do it ex-ante.
  
• It was suggested that the decision maker's policy context (the pressures, constraints, influences, etc.) are a key vehicle to policy influence; probably more than research itself. And based on this, would a stakeholders' context analysis not be the key tool?