Speakers:

Dr Vincent Cable – MP for Twickenham
Julius Court – Research Officer, RPEG, ODI

Chair:

John Young – Programme Manager, RAPID, ODI

1. This was the second meeting in the series 'Does evidence matter?'. John Young introduced the session by pointing to the importance of political context for evidence-based policy. Some of the key questions in this area are: How does the political context affect decisions? What sort of evidence is available to policy makers in different contexts? Are policy makers 'evidence aware'? What other factors influence their decision-making?
   
2. Vincent Cable started off by stating that evidence based policy is important to him as an MP, and emphasised that in many ways researchers and policy makers are in the same business of extracting and processing information.
   
3. He then went on to outline five S's that limit evidence-based decision-making: Speed, Superficiality, Spin, Secrecy and Scientific ignorance.
   
4. Speed: Policy makers are under chronic time pressure and are forced to process information quickly. This requires improvisation and also means that sometimes compromises have to be made. Occasionally, this leads to bad decisions.
   
5. Superficiality: Each policy maker has to cover vast thematic fields, and cannot possibly have in depth knowledge about every issue in those areas. They are therefore heavily dependent on the knowledge and integrity of the people who inform them. This raises difficult questions about who policy makers should turn to for advice, and how they can judge the advice given to them - for example the increasing amount of advice coming from the NGO sector.
   
6. Spin: In the political world, perception is very important. For example, even though evidence has shown that beat policing is not the most cost effective way of using police resources, this form of policing is still prioritised because there is a strong public perception that it will improve security. Perception guides political decisions.
   
7. Secrecy: Vincent also raised the question of how to relate to evidence that is secret. A recent example is Blair's memorandum on weapons of mass destruction in Iraq, which formed the basis of political decisions.
   
8. Scientific ignorance: There is a growing suspicion towards science and scientists among the public, which will have an effect on policies. One example of this is the public demand for zero rail accidents while road accidents are tolerated. This means that political decisions are made to invest far more in rail safety than in road safety.
   
9. Despite the challenges that these five s's present, Vincent concluded by pointing to positive examples where evidence has indeed informed policy, and stated that as policy makers become more professional research will have an increasing role to play in decision making processes.
   
10. Julius Court spoke about the role of the political context in research-policy linkages in developing countries. He introduced the topic by saying that evidence can matter, but that it often does not matter. The question to be asked is therefore: When does evidence matter?
    
11. He presented findings from a synthesis report of 50 case studies on research/policy links in developing countries carried out by the Global Development Network (GDN). The main conclusion from the case studies is that political context is a crucial factor. It is important to bear in mind that the political context in developing countries is distinct due to three factors: (i) Diversity of Southern contexts; (ii) Weak capacity in South; and (iii) Importance of Northern research, influence and funding.
    
12. Five of the GDN case studies were presented briefly: the GALASA case study from India, rainwater harvesting in Tanzania, DELIVERI in Indonesia, animal healthcare in Kenya, and SPEECH in India. More details on these cases can be viewed on the GDN website.
    
13. The political context is clearly the most crucial issue regarding the uptake of evidence both in democratic and less democratic political systems. Evidence does appear to be used more in open political systems, but this depends on the specific issue. Attempts to change the political context usually takes massive effort.
    
14. In general the likelihood of policy uptake can be described using the following formula: Policy uptake = Demand - Contestation, where demand refers to policy maker's and societal demand, and contestation to the degree of variance with prevailing ideology and vested interests.
    
15. Understanding the policy process is crucial for researchers who wish to have an impact. Evidence uptake is greater and also more rapid during crises or policy windows. These windows are hard to trigger, but important to seize.
    
16. A perspective of policy highlights policy implementation rather than formal policy statements. The work of 'street level bureaucrats' who put policy into practice is more visible than the policies themselves.
    
17. Many of the existing theories in the literature on research/policy links are of limited use in a developing country context.
    
18. Strategic factors influencing research uptake include: the level of the policy (macro or local level); current political interests, the political culture, the process (influencing through participation or insider connections); and the importance of timing (responding to decisive moments or longer-term programmes and pilots on the ground).
    
19. Julius concluded by pointing to three areas that need further analysis: (i) the implications of a changing political context within international development policy; (ii) the impact of external influence on policy, for example the impact of donors on national policy processes; and (iii) the consequences of democratic deficits and how to work in less democratic contexts.
    
20. Next steps for the RAPID programme at ODI include: Developing a more systematic understanding about evidence use in different contexts; Developing a taxonomy of contexts; a cross-country study on the evidence/policy links relating to HIV/AIDS; and workshops with NGOs and policymakers.
    
21. Comments focused on a range of issues:

• Would British parliamentarians be better equipped if they had the research facilities provided to the US Senate?
• NGO research does not necessarily lead government astray. Pressure groups have an important role to play in for example select committees that seek a broader and better understanding of different policy issues.
• Policy makers (and not researchers) should take responsibility for distinguishing between high and low level research.
• Vincent's five s's are not constraints that we are forced to work around; they can be changed.
• Policy departments often feel like a different universe to research institutions. Policy makers need evidence that is sufficiently good quality, but cannot afford to wait indefinitely for it. Researchers may therefore have to compromise.
• Working with policy makers requires that researchers make choices about who to approach and when to do so. At what point is it best to engage with the policy cycle?
• The political context is seen as rigid. But from the PRSP case study (conducted as part of the RAPID programme at ODI) we can see that it is possible to change at least the political perception of the context within a relatively short period of time.
• There are differences in political culture between the US and UK, where policy and research are separate, and Germany and Scandinavia, where there are closer connections between the state and research institutes. Is DFID moving towards the model of closer connection? If so, this will have consequences for researchers. They are more likely to have impact on policy if they have an 'insider' status in relation to policy processes than if they have 'outsider' status.
• Who are the policy makers? Should we rather use the term policy actors?