Seven reasons why research doesn’t become policy

Over the last decade there has been increasing demand to make research more useful and applicable for policy-making. And that is also very true within the field of development research. Donors and practitioners need, for instance, to make sure that aid is used in more efficient ways to reduce poverty and inequality—something that the tax-paying public in donor countries expect. At the same time, research funders want value for their money.

A lot of effort has gone in to building theories of social change, establishing methods, and equipping toolboxes. But the task is a demanding one, and our people at UNU-WIDER are working hard on ReCom–Research and Communication on Foreign Aid. Our concern is how development works, which differs from working fordevelopment, since we do not campaign.

Some are better at working for development than others. Recently I had the pleasure to moderate a conference session with Robert Chambers, from the Institute of Development Studies at the University of Sussex, where he spoke strongly for one of the neglected areas in development—sanitation. The sad fact is that 2.5 billion people lack access to toilets, most of them in India. Chambers is a legend in the development field, and has successfully been boosting participatory approaches, policies and actions that promote sustainable livelihoods.

One can question what the policy component is in development as another of the participants, David Mosse from the University of London, has done. His observation from many years of field work is that evidence-based policy is not necessarily influencing development practices. Robert Chambers noted, ‘there is one theory of change for each context’.

Last year we asked participants at a UNU-WIDER conference why research does not always turn into policy. Answers were somewhat mixed, some blaming researchers for sitting on their high horses, others claiming that policy makers are not really interested.

This year we dug deeper, and carried out a whole survey with our ReCom researchers, including in-depth interviews, to find out how they problematize the evidence-based agenda. We also wanted to know how researchers actually work with policy makers to achieve impact.

Every respondent agreed that policy-impact is something they consider already when the research is being designed, even though it might not be an overriding concern. Somewhat surprisingly, and possibly somewhat conversely, the majority said they had not felt pressure to become more policy-relevant.

The answers also point to a number of reasons why research often does not become policy. The list is not new; these are problems that have been spotted in earlier research on the issue, by for instance Bogenschneider and Corbett (2010), or Ray Pawson (2002). Carol Weiss (1995) also pointed to the interaction of the four Is: interests, ideology, information, and institutions.

Based on earlier research and this survey, seven main areas of concern emerged:

(1) the character of the policy-making process

(2) academic systems that do not reward policy-relevant research

(3) complexity of democratic decision-making

(4) limitations of science

(5) social problems are complex

(6) disconnect between the two worlds of researchers and policy makers

(7) measurement of policy impact

The last point is especially problematic. There is no set of metrics to monitor and measure policy impact which makes the task of establishing, let alone proving, influence difficult. Though log frames are often now used they only cover intended effects and not surprises—nor unintended findings, which is one important feature of research.

And there is the time issue. Researchers are working in a different timezone than policy makers who are largely dependent on the electoral four-to-six-year cycle. Two examples from UNU-WIDER: Amartya Sen received the Nobel Prize in Economics in 1998 partly for research he did in Helsinki in the 1980s. Our research on economic inequality some ten years ago might also have partly inspired the Occupy Wall Street-movement. These examples are not something that a log frame could have captured.

Some constraints seem to be eternal: researchers and policy makers are believed to live on two different planets—even in separate universes. There is a lot of talk about building bridges between the two communities, or in this case, sending messengers by spacecraft to bring the evidence to the other world.

And even though a lot of communication work is done to ensure that the message comes through, the question still is: who is listening and in what way? Experiments with policy briefs show, for instance, that only in cases where the receiver has no pre-conceived views on a subject will that person be convinced by new evidence.

The authority of the messenger is important. Our survey shows that researchers believe that they are themselves best positioned to take care of policy-impact and the preferred method is meeting policy makers face-to-face in seminars and conferences. There is also reason to believe that this argument is supported by policy makers. One recurring theme in the feedback we have collected from ReCom results meetings where researchers, aid practitioners and policy makers interact, is that there was too little time for discussion.

But forms of communication other than personal engagement are still crucial because of the huge supply of research and insights that are competing on the market for new ideas. Social media, especially blogs and Twitter, has become increasingly important to get the message out, and not just in the Western world.

Finally, compared with earlier research, this survey also underlines a concern that research funders might underplay. That is the extent to which respondents think that the academic reward system effectively prevents policy-related work. This concern is fuelled by very practical considerations. Academic careers are almost solely advanced based on the number of peer-reviewed publications and citations; unless universities and research funders start changing the system there cannot be a big shift. While academic rankings are used by funders as a key component in the evaluation of research proposals they also need to ensure that policy-related work is appreciated.

Carl-Gustav Lindén is Senior Communications Specialist, UNU-WIDER