Speakers:

John Young

Simon Maxwell

Julius Court

Alan Whiteside

James Putzel

Tony Barnett

Daniel Low-Beer

Douglas Webb

Sue Lucas

Simon Maxwell opened the event and explained that this event aimed to bring together issues around HIV and AIDS, bridging research and policy, and to find ways to do policy better in this very important area of development work. Effectively bridging research on HIV and AIDS, with policy on the disease could make a crucial difference to the trajectory of the epidemic as a whole.

Julius Court then outlined some particular aspects of bridging research and policy on HIV and AIDS. He pointed out that whilst the disease is a very long term problem, which takes several years to become full-blown for individuals, policy cycles addressing the problem tend to be much shorter. HIV and AIDS is a particular challenge for research and policy communities. The problem is extremely complex; it requires an understanding not only of the biomedical aspects of the disease, but also sexual behaviour, politics and culture, the micro and macro environment.

John Young described RAPID's definition of research as a 'systematic effort to increase knowledge' and policy as 'a course of action'. He introduced RAPID's analytic framework for analysing research in the policy process; political context, evidence, links and external factors. He explained how the analytic framework maps onto practical awareness of development work and emphasised that the framework is intended to aid both policy-makers and researchers on their own practical next steps.

Julius Court presented the RAPID programme work on AIDS and HIV. He suggested that HIV/AIDS was a particularly interesting and important case for bridging research and policy, because in some policy areas it seems that the evidence is relatively clear and consistent, but policies remain very varied across the world. As part of its project, RAPID has undertaken an annotated bibliography on HIV/AIDS a literature review on horizontal communication in HIV responses some issues studies and some country studies. Using the RAPID framework, he outlined the central features of bridging HIV/AIDS research and policy, focusing particularly on the case of Uganda.

Questions and comments from the floor addressed a range of issues, including:

• The need to better capture ideas about power within the RAPID framework.
  
• Because HIV/AIDS is such a fast developing epidemic, sometimes research is conducted in one period for a particular purpose, but applied at some other time in a totally different way.
  
• I find the links area of the framework the most unclear - we need to think about links in a way that is culturally sensitive.
  
• We shouldn't assume that external factors like donors are always in the background for bridging research and policy - sometimes they are important and sometimes they are not.
  
• In much of Africa there is still a deep suspicion about Western motives. Often, the central aspect of how research can fit into policy is where research is seen as coming from.

Alan Whiteside opened his presentation by pointing out some of the difficulties with HIV research; there are numerous gaps in the data. Even mortality statistics are often contested and denied, and it is very difficult to track spread of such a fast developing disease in countries with poor research resources. Alan outlined the key challenge: how to convince people to change their behaviour and do things they didn't automatically want to do. There are some very difficult policy choices to be made by policymakers. The key issue is that policymakers need to understand why they are making their choices. For example, some interventions have a very high economic cost per life saved and policy makers can raise legitimate questions about whether this is always the best way to direct resources.

James Putzel introduced the findings of his recent study on decision making processes in Uganda and Senegal (for details. Two main sets of issues were highlighted as particularly important:

• political leadership: The incentive structures for both policy makers and researchers are very important, along with the political legitimacy of those making decisions. Leaders need to have more to gain than to lose in taking up fight on HIV and AIDS.
  
• the organisation of political authority: the application of multi-sectoral approaches, decentralisation, and any local dependency on external resources can be very influential. Many organisations are not based on any research at all - but the key is how leadership positions itself in relation to evidence. This should be the key preoccupation of bridging research and policy; it must become political illegitimate not to engage with HIV aids.

Questions and comments from the floor:

• Often HIV and AIDS are very politicised - it can be difficult for politicians to act even if they are persuaded by researchers.
  
• The more dissent you have within an organisation on a particular point, the more robust your evidence needs to be. The politics within an organisation dictates the kind of evidence that needs to be mustered.
  
• Many political decisions have nothing to do with evidence, and don't really need to be based on evidence in order to be good decisions.
  
• We should be working to support Southern research capacity and think about how this Southern research can feed into Southern policymaking.
  
• Should we go for political prominence by campaigning on a very limited area or try for a more holistic approach to HIV/AIDS policy?
  

Tony Barnett focused on the use of evidence and knowledge within DFID's work on HIV and AIDS. After DFID was made aware of HIV as issue in 1980s, evidence was not used in the setting of the strategy. The 2001 HIV advisory strategy paper was a corporate response trying to revamp current approaches to the problem internally. The major organisational constraint that prevents getting research findings into policy in DFID is lack of time. Also, the structures of DFID themselves have prevented multi-sectoral responses and therefore obstructed the path of research into policy. DFID has struggled to learn from its own research.

Daniel Low-Beer suggested that there should be a creative tension between research evidence and policy making. It is crucial to have primary evidence in research, and that this evidence is in the public domain. This facilitates public discussion and allows research to be an effective process; not just produce useful products. It is important to keep our eye on the ball and remember who research is for; it can be for policy makers, but it can also be for the general population. He also noted that it has been hard to understand HIV decline in Uganda because the data doesn't fit policies very well. Casual sex has decreased around 65% in recent years, but we are not sure why. Population response to the epidemic is hard to quantify; we need to ensure we have sufficient qualitative data to understand these shifts.

Questions and comments from the floor:

• The problem much of the time is that there just isn't enough good research around.
  
• The way that DFID uses knowledge is still fairly ad hoc.
  
• Multi-sectoral work is very hard, and it is very expensive.
  
• At what point is evidence, evidence? It is difficult to tell when you are working from an NGO perspective and you have an explicit vested interest.
  

Douglas Webb first outlined the background to bridging issues on HIV and AIDs; pointing out that many in the NGO sector have gone from a service delivery role to a more policy entrepreneurial activities. In this context, it may well be that some NGO research is best directed at the media, with a view to creating media coverage which will influence policy, rather than seeing policy makers as the primary audience to NGO research. If we are undertaking advocacy in the policy process, we need to think about how we can establish our legitimacy. We need to have a credible mandate; to represent a particular interest group or constituency. Often, establishing this constituency is a very fraught and stressful process. There is a technical problem with evidence in the NGO section - time frames are often too short to generate rigorous and credible research. We need to think about where we really want to have policy dialogue - in many senses, country level is the most important in many respects, because this is where practical change is really played out.

Sue Lucas suggested that the key question in this discussion was 'whose policy are we talking about?' From her experience in Thailand and Uganda, she identified three questions: (i) Is there different research we should be doing? (ii) If so who would be the key partners? (iii) Should there be changes in the relationships between the partners? She noted that the current research agenda is dominated by interventions - by looking back at what has happened before. Often, this is not the most relevant information when thinking about how we should respond next. She emphasize the needs to make links between policy and research communities more immediate and for developing direct links between researchers and communities. Links tend to be effective in different ways in different countries.

Questions and comments from the floor addressed a number of issues:

• We have known for years that certain strategies do work - there is concern that initiatives like 3 by 5 will draw the focus away from prevention.
  
• I am surprised by the levels of consensus here today on the harmful effects of 3 by 5. But, sometimes it is simply unfeasible to wait for the evidence before we act.
  
• The media has a big role to play in persuading policy makers, and it may be that getting research into the media is the best way to affect political processes.
  
• This disease is in many ways a humanitarian crisis, but we should keep our heads and focus on building local capacity to cope, and hold on to the use of evidence for as long as pragmatism will allow.
  

The final session called for suggestions for next steps and asked the participants to indicate what they kind of future work would be useful on bridging research and policy on HIV/AIDs.

Julius Court introduced the session, explaining RAPID's work on HIV/AIDS thus far has been a scoping exercise. In the future RAPID needs to:

• Do deeper country case studies, perhaps including a wider set of issues and focusing on specific evidence cases.
  
• Engage CSOs who are working in HIV policy and the use of evidence in action research.
  
• Do more systematic, cross-country work.

What can ODI do to help bridge research and policy gaps on HIV/AIDS more quickly? Contributions from the floor included:

• It might be interesting to look at different institutional responses to the same issue - and do a comparison of organisations across the world dealing with the same problem.
  
• You could do a case study of a global policy process, like 3 by 5 or the setting up of the global fund. The idea of case studies is really important.
  
• Overall, RAPID should take on an action-research mindset. ODI has the opportunity to bolster evidence that often struggles to find credibility; it could be very valuable for improving HIV/AIDS policy to validate the kind of evidence you gather working with non-researchers and using qualitative methods.
  
• It is important to engage with the scientific methodological process in this work. There are particular difficulties when answering policy questions from the point of view of scientific research.
  
• It might be interesting to do work on prevention; perhaps you could look at the failure of condom policy over the last 20 years.
  
• You need to be clear about whose prerogative it is to make policy. We need to think about how to build that capacity to engage those who are genuinely from the South.