ROB STEPHENSON AND MONIQUE HENNINK

MovingBeyondResearch toInformPolicy:Barriers and Strategies in Developing Countries

November 2002

Contents1 Authors

1 Acknowledgements

2 Summary

3 Introduction

4 The Process of Knowledge Transfer

5 Methodology

5 Current Methods of Dissemination and Uptake of Research Results

6 Researcher’s Barriers to Dissemination of Research OutputsPolicymakers Perceptions of ResearchEmphasis on StatisticsLack of Dissemination Skills and Access to PolicymakersLack of ResourcesDonor Influences

9 Policymaker’s Barriers to Uptake of Research OutputsLimited Access to Research OutputsLack of Central Source of Research OutputsQuality of Research

10 Mutual Barriers to CommunicationLack of Formal Communication ChannelsLack of Collaborative ResearchFormat of Research OutputsPolitical Influences

13 Discussion

16 Conclusion

17 References

AuthorsRob Stephenson Ph.D. was a Research Fellow in theOpportunities and Choices programme of reproductive healthresearch at the University of Southampton, UK, at the time of thiswork.

Monique Hennink Ph.D. is a Senior Research Fellow in theOpportunities and Choices programme of reproductive healthresearch at the University of Southampton, UK.

AcknowledgementsThis work was completed under a Small Grant from theOpportunities and Choices Reproductive Health ResearchProgramme, University of Southampton, UK, funded byDepartment for International Development. The authors wouldlike to thank all those who contributed to this research in thestudy countries.

Suggested Reference:

Stephenson, R and Hennink, M (2002) Moving Beyond Researchto Influence Policy: Barriers and Strategies for DevelopingCountries. Opportunities and Choices Working Paper No 2002/05Univeristy of Southampton, UK.

Contact:Opportunities and ChoicesReproductive Health Research ProgrammeDepartment Social StatisticsUniversity of SouthamptonSouthampton SO17 1BJTel: 023 8059 2527Fax: 023 8059 3846www.socstats.soton.ac.uk/choices

1

SummaryThis paper examines issues of communication across theresearcher-policymaker interface in four countries: Malawi,Tanzania, India and Pakistan.

In-depth interviews were conducted with researchers working inhealth issues, and policymakers responsible for the formation ofhealth policies at the local and national levels. The range ofstrategies used by researchers to disseminate research outputs topolicymakers was similar across all study countries: mostresearchers disseminated their findings through research reportsand workshops, and there was a heavy reliance on disseminationthrough academic circles.

Policymakers reported difficulties in accessing research outputs,particularly when disseminated through academic channels. Theinterviews identified a number of barriers to effectivecommunication between researchers and policymakers. The lackof clear communication channels between researchers andpolicymakers and the lack of a central depository for researchoutputs restricts the dissemination of research outputs.

Researchers felt that policymaker’s lack of understanding andrespect for research limited the extent to which research is usedin policy formation, whilst their own lack of skills and resources fordissemination restricts the effectiveness of their disseminationefforts.

Policymakers felt that the research outputs they receive often lackpolicy recommendations, were of poor quality and werepresented in academic formats.

The results of this study highlight a communication gap betweenresearchers and policymakers. Whilst reducing this gap requireslong-term attitudinal change and an increase in investment in theresearch sector in these countries, other barriers are more easilysurmountable.

Training in communication skills is necessary to allow researchersto target their research towards a policy audience, whilstpolicymakers need to take steps to inform researchers of theirinformation needs.

A greater understanding of the potential contribution of researchto policy and the constraints of policy formation could arise fromcloser collaboration between researchers and policymakers.

2

Introduction Research is generally understood to be a systematic process forgenerating new knowledge, and can act as a powerful tool forproviding information for policy formation (Walt 1994).

The continuing trend towards evidence-based policy formationhas increased the demand for research outputs that can provideclear, concise policy-relevant findings (WHO 2000: ECDPM2000). A prerequisite for evidence-based policy formation is thetimely provision of scientifically solid and up-to-date informationto policymakers (WHO 2000: Bark 1979). Indeed, it has beensuggested that the success of a country’s development effortsnow depend upon the degree to which its planners and programmanagers use and apply research for decision-making (Torres1981). Demographic research continues to produce a large bodyof scientific findings that can address critical issues faced bypolicymakers, and informed policymaking can benefit fromunderstanding the policy implications of such research (RAND1997).The extent to which such research is translated into policyaction, however, is dependent on the success of communicatingresearch outputs between researchers and policymakers.

The effective dissemination of research results to policymakers is anessential element of any research programme, not only as a meansof translating research results into policy action, but also to provide‘pay-back’ for the investment in health research (Askew 2002).

The growth of large donor-funded operations researchprogrammes focussed towards developing countries hasincreased the need to provide evidence of policy impact, which inturn has placed greater importance on ensuring that researchoutputs are communicated effectively to policymakers.

Effective communication between researchers and policymakers,however, provides a continual challenge. Previous research hasidentified the existence of a gap between the amount of researchthat is produced and the amount that is implemented throughpolicy changes, resulting in the current under-utilisation of healthresearch (Walt 1994b: RAND 1997).

This paper examines issues of communication across theresearcher-policymaker interface in four countries: Malawi,Tanzania, India and Pakistan. Although previous research hasidentified the existence of a communication gap betweenresearchers and policymakers (Walt 1994:1994b), little is knownof the factors creating this divide. This paper explores currentmodes of research dissemination between researchers andpolicymakers, and examines the barriers to effectivecommunication between the two parties. An understanding ofthe difficulties faced by researchers and policymakers indisseminating and receiving research outputs has the potential toincrease our knowledge of the communication process, and tohighlight strategies for the effective dissemination of researchoutputs.

3

The Process of Knowledge Transfer

The study of the researcher-policymaker nexus can providevaluable lessons for increasing the effective dissemination ofresearch outputs (Walt 1994). In order to understand the barriersfaced in communication between researchers and policymakers, itis necessary to understand the process through which suchcommunication occurs. Weiss (1977) suggests that there are twoalternative models to describe communication betweenresearchers and policy-makers.The first model, the ideal model ofthe natural sciences, describes research as a linear series of events,which cumulates in the dissemination of research outputs. Themodel postulates that the research process is a logical one whichfollows a particular sequence in which researchers are assumed toask the right questions, plan and conduct research rigorously, andthen disseminate the findings (Walt 1994:Weiss 1977).

In this model communication between researchers andpolicymakers provides the end-stage in the research process, withresearch outputs passed directly to the policymakers. Thealternative approach, the enlightenment model, argues that the linksbetween research and policy are less direct, and that no single pieceof research is likely to influence policy change directly (Weiss 1977).Rather, it is suggested that research influences policy throughproviding a cumulative weight of information, which leads to agradual change in the thinking of policymakers. Research outputspermeate gradually into the policy process through a number ofinformation channels, providing a backdrop of information thatinfluence policy (Walt 1994). Smith (1993) argues that this modelprovides a more realistic view of the researcher-policymakerinterface, noting that the exchange of information betweenresearchers and policymakers is a dynamic process involving arange of channels and actors.

The uptake of research outputs can be either deliberate or diffusive(Walt 1994). Deliberate dissemination occurs when policymakersactively seek new information from researchers in order to answerspecific policy questions. In contrast, diffusive dissemination occurswhen policymakers learn of research findings through a range ofresearch outputs, and knowledge is gradually diffused topolicymakers, as in the enlightenment model of communication.Porter and Prysor-Jones (1997) argue that deliberate disseminationis most likely to result in policy change as the stimulus to seekinformation arises directly from the need to provide information fora specific policy problem.Weiss (1991) notes that there are threeforms of research dissemination: research as data and findings,research as ideas and criticism, and research as briefs andarguments for action.The extent to which research influences policyvaries between the three types of research outputs. In the first,research as data and findings, research is communicated topolicymakers in the form of statistics and academic reports.Weiss(1991) argues that policymakers will only use this form of output if

4

specific problems exist which can be answered by the statistics orif they highlight an immediate and acute problem. The second,research as ideas and criticism, involves the gradual flow ofinformation between researchers and policymakers, as described inthe enlightenment model. Issues arising from the research graduallypermeate the policy process through a variety of channels, and it isdifficult to identify the specific piece of research that leads to policychange. Finally, research as briefs and arguments for action involvesresearchers adopting the role of advocates. Such research oftenprovides the policymaker with a set of alternatives to a currentsituation, and is therefore most commonly utilised by policymakers.

MethodologyThis research was conducted in four countries: Malawi, Tanzania,Pakistan and India, to facilitate cross-national comparisons of theissues under investigation. Data were collected via in-depthinterviews with health researchers and policymakers responsiblefor developing health policy. Independent researchers and thoseaffiliated with research institutions were identified, and theprinciple researcher in health issues was invited to participate inan interview. Policymakers were identified as those having a directresponsibility for the formation of health policy at the provincialor national level in each country.A total of 48 in-depth interviewswere completed.

Data were collected using a semi-structured discussion guide.Respondents were asked about their current practice of researchdissemination and uptake; their opinions on the effectiveness ofcurrent dissemination strategies, barriers to the dissemination ofresearch; and suggested strategies for improved dissemination.Theuse of semi-structured in-depth interviews allowed the collectionof reliable comparable data from respondents, whilst still allowingthem to express beliefs and opinions in their own terms (Bernard,1994; Patton, 1990). All interviews were tape-recorded and latertranscribed. Textual data analysis was conducted using thematicanalysis and comparison between countries to identify nationalpatterns.

Current Methodsof Dissemination

and Uptake ofResearch Results

The range of strategies used by researchers to disseminateresearch outputs to policymakers is similar across all studycountries. Differences in the dissemination process, however, existbetween commissioned (by a donor or a government agency)and non-commissioned research. For commissioned researchthere is a direct channel of communication between theresearcher and the commissioning agency, facilitating thedissemination of the final research outputs. A commissioning

5

agency is typically involved in various stages of the researchprocess and has a vested interest in the research outputs thatusually take the form of a research report. Additionally,commissioned research is often disseminated via donor-fundedworkshops, conducted with a range of in-country stakeholders, orthe donor agency initiates the distribution of the research outputsto a wider audience.The research outputs most frequently usedby policymakers were those from internally commissionedresearch.

For non-commissioned research the channels of dissemination topolicymakers are less clear and more varied.Many researchers limiteddissemination to academic channels (e.g. papers in peer-reviewedjournals or presentations at research conferences). The directdissemination of non-commissioned research to policymakers mostcommonly involved either the distribution of a research report to arange of policymakers or inviting key policymakers and otherstakeholders to a dissemination workshop.

Policymakers reported seeking research findings only when they hadspecific information needs. If the information is not available internallyor through commissioned research outputs, policymakers consulteda range of sources including other ministries and governmentdepartments and documents from international researchorganisations. To a lesser extent, policymakers contacted universitydepartments and national research organisations, however, thischannel was only used if there was an established link with theorganisation. In many cases consultants were employed to locaterelevant published material or to conduct a research study.

Both researchers and policymakers reported that there exists nostructured mechanisms through which dissemination can take placeeffectively, this is highlighted by the disparate strategies ofdissemination used by researchers and the sources of uptake ofresearch used by policymakers.However, for commissioned research,there is a direct channel of communication between researchers andpolicymakers.

Researcher’s Barriers to

Dissemination of Research Outputs

Policymaker’s Perceptions of ResearchResearchers reported that the policy environment in the studycountries is not focussed towards the incorporation of research inpolicy formation and program development. This lack of a strongevidence-based culture in policy development was felt to be asignificant barrier to the dissemination of research to policymakers.Researchers perceived that research is given low priority bypolicymakers and that research findings are not valued in policyformation.As a result academic research is often seen to have littlecontribution to the policy development process, and policymakers

6

are seen as not fully appreciating the potential contribution ofresearch in enhancing policy formation. In addition, researchers feltthat research is perceived as an unnecessary expenditure for policydevelopment in resource poor countries. For example;

The culture of using research results for policy does not exist inTanzania (researcher,Tanzania).

There is a general feeling among policymakers that as far aspolicymaking goes they are the experts. If you want to bring inresearchers they are just there to punch in numbers (researcher,Pakistan).

Policymakers don’t see the role that research plays in everydaysituations. Senior government officials don’t appreciate the role ofresearch in programmes (researcher, Malawi).

The ministers make the policies themselves, without using what wesend them, they don’t realise that research could help them(researcher, India)

Researchers also suggested that policymakers might not fullyunderstand how to use research to support policy formation.They noted that policymakers might not have the ability toevaluate the quality of a research study or to interpret researchfindings, thus experiencing difficulties in translating researchfindings into policy action. This may lead to the failure toincorporate research into policy or to extraneous conclusionsdrawn from research results.

They do not understand the research process, so they do notdifferentiate between good and bad research...many of them don’teven understand the difference between qualitative and quantitativeresearch (researcher, Pakistan)

Policymakers may not understand research.They feel that it may takeyears to get an outcome, they are interested in outputs for today.There might be a lack of vision on the part of the policymaker(researcher,Tanzania)

Emphasis on Statistics Researchers reported that policymakers place a strong emphasison the use of statistics from research outputs. While researchersrecognised the importance of measurement indicators, and theincreasing pressure on policymakers to justify programmes orquantify changes, they felt that the focus on statistics under-utilisedresearch results. Researchers felt that substantive issues arisingfrom the research could also impact on policy. Policymakersthemselves acknowledged their focus on statistical informationfrom research outputs.

They are interested in a few indicators, for example, what is the CPR?Which they have to report to their highers, but other areas that arereally important such as quality and side effects are not given asmuch attention (researcher, Pakistan).

Basically it’s statistical information we require from the researcher

7

because they have it readily available and can provide it promptly(policymaker, Pakistan).

Lack of Dissemination Skills and Access to PolicymakersMany researchers identified their own lack skills in dissemination ofresearch to policy audiences as a key barrier to the successfuldissemination of their research outputs. Researchers highlightedtheir lack of training in communication of research findings outsideacademic circles. They were also uncertain about whom todisseminate research findings to, and expressed difficulties inidentifying and accessing policymakers. Although some researcherswere able to identify policymakers, they reported that the frequentchanges in government portfolios meant that developing andmaintaining links with policymakers was problematic.

Lack of ResourcesAnother fundamental barrier is the lack of resources fordissemination activities, particularly in Malawi and Tanzania, whereresearchers reported that often there are no funds available toconsider dissemination activities beyond the distribution of aresearch report. Whilst resource issues were less apparent fordonor-funded research, for Government sponsored orindependent research, resource constraints often curbeddissemination activities. The lack of resources also restricted theprofessional presentation of information for policy audiences (e.g.brochures or fact-sheets).

How dissemination works will depend on the amount of resources,resources for dissemination are limited, so you can’t disseminate in theway you would like to due to shortages, we don’t have PowerPointpackages also to present findings nicely (researcher, Malawi).

Donor Influences Much health research in the study countries, particularly in Malawiand Tanzania, is initiated and funded by international agencies. Theprominence of donor-driven research raised a number of issuesrelated to dissemination. First, researchers highlighted the commondisparity between the health issues investigated by internationaldonor agencies and the priority areas of national policymakers, hencepolicymakers often disregard research outputs that are not focussedon national areas of priority, for example;

Most research is donor-driven, this may not be of interest to theMinistry. Research, which is policy oriented, is usually donor-driven, soseen as not relevant (researcher,Tanzania)

Researchers are sometimes forced to conduct research on which wehave no interest, but are instructed by the donors. All these thingsfrustrate the policymaker at the end (researcher, Malawi)

Second, researchers are typically involved in donor-initiated researchas consultants and are not obliged to become involved in

8

dissemination activities. Thus, a protocol has developed wherebyresearchers pass research findings only to the donor agency and thenmove on to the next consultancy. In this situation, the onus is on thedonor agency to disseminate research outputs more widely.Researchers felt that involving researchers in the disseminationprocess would be beneficial as they were more aware of the contextof the issues brought out in the research.

Third, research conducted by international agencies based outsidethe study countries often has limited time to consult nationalpolicymakers and identify priority research issues. Researchers feltthat this situation leads to a limited understanding of the culturaland contextual issues surrounding the research, and that therewere particular limitations with policy recommendations that failto reflect ‘on the ground’ realities.The research outputs were alsomost likely to be presented at international conferences orpublished in international journals, and hence not easily accessibleto local policymakers and government ministries.

Research is donor-driven rather than created through need. Thepolicymakers don’t take these research recommendationsseriously..external consultants also are not aware of Malawi culture andso annoy those who are responding to the research (researcher,Malawi).

Most research ends up on the shelves of people abroad and can’t beaccessed by policymakers in Tanzania (researcher,Tanzania).

International agencies give a background but local agencies give thecontext of India. International agencies should have liaison with Indianorganizations to understand the complexities on the ground(policymaker, India).

Policymaker’s Barriers to Uptake of Research Outputs

Limited Access to Research Outputs Accessing research outputs is one of the greatest difficulties facedby policymakers in utilising research findings. Policymakers identifiedthat research outputs often do not reach ministries or governmentofficials as they are disseminated in academic circles or passed onlyto donor agencies. Policymakers seldom access these channels ofdissemination, so scientific information with the potential tosupport policy formation has limited accessibility to all the relevantstakeholders. Policymakers reported a need for wider in-countrydissemination of research outputs.

The research that has been conducted is usually by the academics orthe universities, and is published in the international journals and sothey don’t get shared at the local level or the country level(policymaker, Pakistan).

Research just goes to those who commissioned it, not disseminatedwidely information flow is usually unidirectional, it does not go furtherhorizontally (policymaker, Malawi).

9

Researchers do research for dissemination in publications or technicalpapers where policymakers don’t get the chance to read.There is stilla gap in utilising the available research into programming orpolicymaking (policymaker,Tanzania).

Lack of Central Source of Research Outputs An issue identified mostly by policymakers in Malawi and Tanzania isthe lack of a central depository for health research outputs.Policymakers expressed frustration at trying to locate researchreports without a central clearing-house or database where theycould easily access health research.A depository of research outputswas thought to be particularly useful for accessing unpublishedstudies. It was suggested that such a depository be co-ordinated bya public institution to enable greatest access and should bepublicised widely among the policy community.

Quality of Research The quality of in-country research was an issue that discouragedsome policymakers from using local research outputs and promptedthem to seek research findings from international agencies.This wasa particular issue in Malawi and Tanzania, where policymakersidentified that the small skill base amongst local researchers made itdifficult to conduct high quality research. Policymakers felt that thelack of government investment in the research sector meant thatlocal researchers have not had the advantage of expert trainingafforded to those in international agencies, and so the technicalcompetence to undertake research of an international standard wasnot fully developed.

There are quality issues with in-country research, greater validity isgiven to research conducted by international agencies, as long as theyused local people to field it (policymaker, Malawi).

The difference in quality of local based research versus internationallyconducted research is important. Malawi based research does nothave the advantage of technology to present results in a digestibleform. Even substantive quality is more traditional scientific enquiryrather than programmers needs. International NGOs have workedhard at developing ways to approach policymakers (policymaker,Malawi).

Mutual Barriers to Communication

Lack of Formal Communication ChannelsBoth researchers and policymakers identified the lack of formalchannels of communication as a barrier to effective dissemination.Policymakers expressed difficulties in identifying researchersbeyond informal contacts, and researchers faced problems inidentifying policymakers to whom to disseminate researchoutputs. Both groups felt that a communication forum (such as aresearch association or regular meetings) would enable effectivedialogue between researchers and policymakers. To foster more

10

effective communication between researchers and policymakers,it was suggested that researchers be included on Ministerialadvisory boards to identify priority research areas and defineappropriate research agendas to allow research activities torespond to programmatic needs.

Lack of Collaborative Research Policymakers reported that when they were involved in theresearch process, had commissioned the research, or the researchwas in direct response to a policy need, it was more likely to beutilised for policy development. Researchers also reported thatthe involvement of policymakers in the research process led to amore effective consideration of policy issues, political limitationsand practical realities in implementing the research findings. Bothparties, however, reported that much research was conductedwithout collaboration between researchers and policymakers, andthis posed a significant barrier to dissemination. Collaborationbetween researchers and policymakers was encouraged atvarious stages in the research process; in defining researchproposals, designing research questions, and particularly in shapingpolicy recommendations that are realistic and relevant to theresource constraints of the ministries. Ensuring that policymakersgain a sense of ownership of the research is seen as crucial to theuptake of findings, illustrating the importance of developing arelationship of trust between researchers and the policycommunity.

There needs to be a whole dialogue between policymakers andresearchers at the beginning of the research study, so that it becomessomething that programmers have a vested interest in andresearchers understand that vested interest and try to meet it. Thatmight help to facilitate the uptake of research findings in decision-making (policymaker, Malawi)

As a programmer, if I am involved in the research I will be able tounderstand it better, similarly if a researcher can participate in thepolicymaking level then he will be able to make an impact in a betterway (policymaker, India)

Format of Research Outputs Many policymakers reported difficulties with the format and style inwhich research outputs were presented, stating that research reportswere often written in an academic style using technical language,andinclude complex statistics that are difficult to understand. Policymakersstated that research outputs were often too lengthy and thatconcise, well-structured reports with an executive summary of thekey findings and policy implications were more appropriate.Researchers, however, felt that summarising research findings riskedloosing the detail needed to fully understand the research problem.They also felt that their academic peers would poorly regardresearch if it did not provide adequate detail of the researchmethodology, statistical techniques and use the correct terminology.

11

One of the critical issues for policymakers in using researchoutputs is the frequent absence of policy implications orrecommended interventions from the research. Furthermore,policy implications that are presented are often too general orunrealistic in terms of resources. Some policymakers felt that arange of policy recommendations should be provided such asshort, medium and long-term strategies and that options shouldbe given for various resource scenarios. They also felt thatresearch reports should highlight which agencies should beresponsible for initiating changes. In response, researchers felt thatthey are often not aware of policymakers’ priorities and resourceconstraints and therefore find it difficult to develop feasible policyrecommendations. Researchers also stated that policymakersoften judge the policy recommendations simply on practicalityand affordability of implementation rather than on theimportance of the issue.

It is the how part, how you can change things, what you should do.Researchers usually don’t do that, they put the research on the tableand say now you figure out what to do (policymaker, Pakistan).

They need to interpret figures and findings, not just present the findingthat one third of children born to teenage mothers die – say that thismeans that there should be pregnancy prevention available(policymaker, Malawi)

Sometimes researchers don’t know who to address theirrecommendations to – they say policymakers – but that is a broadterm (policymaker in India).

Political Influences Researchers and policymakers acknowledged that policyformation is often influenced by political priorities andconstrained by the resources of government. It was recognisedthat even well developed research findings may not be actedupon if the political climate was not conducive to change. Someresearchers were reluctant to disseminate research findings thathave political implications and felt that it would be inappropriateto disseminate findings that were in conflict with current nationalpolitics. While these issues did not discourage researchers fromdisseminating research outputs to policymakers, researchers wereaware that complex political issues often influenced policydevelopment.

Everyone who read the report thought it was excellent, the Ministerswere very happy with it, but no one could implement it as it was toopolitically contentious (researcher, Pakistan)

Research findings may be beyond the capabilities of policymakers,even though the problem is highlighted by research. Our researchfound that many women were delivering at home due to lack oftrained staff at hospitals. But policymakers are not permitted toemploy more staff (researcher,Tanzania)

12

DiscussionAlthough this study was conducted in four very different countriesit is worthy to note the uniformity of issues raised by researchersand policymakers, indicating the broadly common experiences inthe process of research dissemination. A number of issues,however, were more prominent in the African study countries, inparticular the influence of donors on the process of dissemination.Also the resource and infrastructure limitations and the smallerpool of skilled professionals in health research in Malawi andTanzania lead to the greater prominence of issues such as the lackof communication networks, the need for central depositories ofresearch information and the lack of in-country expertise forpolicy research in these countries.

The current processes of dissemination between researchers andpolicymakers in the study countries essentially follows the idealmodel of the natural sciences (Weiss 1977), in which disseminationforms the end-stage of the research process. Researchersfrequently reported that policymakers were not usually involved inthe research process, and that they disseminated their findings viareports sent to policymakers as the final stage in their researchprocess.There was little evidence of communication in the studycountries that followed the enlightenment model (Weiss 1977),which reflects the limited channels of communication betweenresearchers and policymakers. An increase in the number andtypes of dissemination strategies used by researchers would benecessary to increase policymakers exposure to research outputs,and to foster the gradual permeation of research into the policyformation process, as described in the enlightenment model. Inaddition, policymakers described a ‘deliberate’ form of researchuptake, in which they sought specific research information orstatistics to support a policy. Diffusive dissemination, which requiresthat policymakers have exposure to research findings from a rangeof sources, was not evident as the channels of researchdissemination were limited in each of the study countries. Thedissemination of research in this study was also strongly focussedtowards the use of research as ‘data and findings’, rather than as‘ideas or criticism’, or ‘arguments for action’ (Weiss 1991); inparticular this was reflected in the frequent absence of policyimplications in research outputs.

This research identified a number of barriers in the study countriesthat inhibit the effective dissemination of research outputs. Acrucial barrier to the use of research in policy formation is theabsence of a strong evidence-based policy culture. Currently thepolicy environment in the study countries is not focussed towardsincorporating research in policy formation, so policymakers do notvalue the contribution of research in the policy process. In addition,researchers often neglect to identify the policy implications ofresearch studies or identify a course of action for policy audiences.Hence, research is either ignored, or policymakers developed theirown extraneous implications from the research.

13

The lack of collaboration between researchers and policymakersin the study countries also acts to prevent effective disseminationof research outputs. Porter and Prysor-Jones (1997) argue thatone of the essential elements for effective communicationbetween researchers and policymakers is the involvement ofpolicymakers in each stage of the research process. They arguethat this fosters a sense of ownership of the research bypolicymakers and ensures that the research meets the specificneeds of the policymaker. Such collaboration is also effective inensuring that the format of research outputs meets therequirements of policymakers.The continued reporting of a lackof policy implications and inappropriate formats of researchreports as barriers to the use of research in policy formationpoints to the need for greater collaboration between researchersand policymakers at all stages of the research process.

Worral (1972) states that developing countries often lack thecommunications infrastructure to allow the effectivedissemination of research outputs, in particular, the lack of aformal system of communication between researchers andpolicymakers, with an absence of "clearing houses" that can act asa store for research information and materials. Hence, existingresearch is not always available to those who need it, andpolicymakers must thus make decisions based on inadequateinformation (White 1993). The absence of a strongcommunications infrastructure proved to be a major barrier tothe effective dissemination of research outputs in each of the fourstudy countries, limiting the channels of communication betweenresearchers and policymakers and restricting communication tothe form described in the ideal model of the natural sciences.Although the ideal model of the natural sciences does result inthe dissemination of research to policymakers, dissemination asthe end-stage of the research process misses opportunities forpolicymakers to become involved in the research process, and toensure that the research meets their information needs. Theindirect nature of the enlightenment model of communication,whereby research information gradually permeates into the policyprocess through a number of information channels is currentlyunderdeveloped in the study countries. Research can have thegreatest impact on policy when effective communication existsbetween researchers, policymakers and the community affectedby policy change (Porter and Prysor-Jones 1997). However, muchresearch communication in the study countries takes placethrough academic channels and does not reach policy audiences.It also risks appearing esoteric and loosing its relevance topolicymakers.

The dissemination of research to policymakers does notnecessarily ensure that it is used in policy formation.Walt (1994)argues that political environments are not always conducive to theincorporation of research into policy formation, especially if theresearch outputs question policy impetus or political ideologies. Inaddition, research that advocates change that may disrupt long-

14

standing power-relationships and organisational cultures takes agreat deal of effort to implement, and as such may be ignored bypolicymakers (Haaga and Maru 1996). Researchers andpolicymakers in the four countries acknowledged the role ofpolitics in policy formation, and noted this often superseded theimportance of research in the policy formation process.

Dissemination between researchers and policymakers is notalways a simple two-sided communication, there may be a rangemediators in the relationship (Askew 2002). Mediators mayinclude advocates, communications experts and donors, whopromote the main findings of research to policymakers, in anattempt to focus policymaker’s attention towards a specific areaof need. Askew et al (2002) argue that the increasing importanceof mediators in the dissemination process has evolved both fromresearcher’s lack of understanding of the policy process andpolicymakers difficulties in understanding research outputs.Mediators thus act as a mechanism for bridging thiscommunication gap, promoting the policy relevance of researchoutputs.The only type of mediators identified in this study weredonors, who often commissioned research and then disseminatedthe research results, using a wider range of disseminationstrategies than were possible for non-commissioned research.Theuse of mediators thus offers the potential to increasepolicymaker’s access to research outputs particularly in settingswith poorly developed communication channels betweenresearchers and policymakers.There were, however, a number ofproblems associated with the use of donors as mediators thatneed to be addressed if donors are to become effectivemediators. These include the potential loss of context in theresearch issues and the risk that research outputs would not bedisseminated in the country of the research.

Researchers themselves identified their lack of dissemination skillsas a barrier to the successful communication of research outputsto policymakers, whilst research outputs were often criticised bypolicymakers for been too complex and lengthy and lacking inpolicy recommendations. These barriers point to the need fortraining on dissemination skills among researchers in the studycountries; to foster a greater understanding of how to prepareresearch outputs for a policy audience and increase confidence inlocally produced research. Conversely, researchers often reporteda lack of understanding of research among policymakers.To someextent this could be improved through increased collaborationbetween researchers and policymakers throughout the researchprocess. This may act to improve policymakers understanding ofresearch outputs, whilst also creating a sense of ownership of theresearch among policymakers.

15

ConclusionThe barriers identified in the process of communication betweenresearchers and policymakers in the four countries weresurprisingly similar, suggesting not only that communicationbetween researchers and policymakers is a widespread problem,but that the divide between the two parties is shaped bycommon factors. The lack of clear communication channelsbetween researchers and policymakers and the lack of centraldepositories for research outputs restricted the successfuldissemination of research outputs, and limited communication tothe type described by the ideal model of the natural sciences. Forcommunication between researchers and policymakers todevelop into the form described by the enlightenment modelrequires greater investment in the research and communicationsectors, to facilitate the growth of a range of communicationchannels and increase the exposure of policymakers to researchfindings. Given the resource poor environments of the studycountries and the range of demands on limited budgets, such aninvestment is unlikely in the immediate future. Mediators,however, provide a potential opportunity for increasing theeffectiveness of research dissemination, and international donorshave a clear role to play in fostering the growth of evidence-basedcultures in these countries. Such mediators, however, must takesteps to ensure that the research they commission and thefindings they promote are contextually appropriate, which can beachieved through increased collaboration with in-countryresearchers and policymakers.

Other barriers identified in this research are more easilysurmountable. Researchers can take steps to improve the formatof their research outputs and to include relevant policyrecommendations, whilst policymakers can communicate theirinformation needs to researchers. Dissemination was oftenviewed as the final stage in the research process, with researchersand policymakers meeting only at this stage. Dissemination,however, is a continual process, with contact between researchersand policymakers at various stages of the research processoffering the chance for both parties to have inputs into theresearch and ensure that research meets current programmaticneeds. Increased collaboration between researchers andpolicymakers was consistently reported as the key to thesuccessful dissemination of research outputs, and requires effortfrom both parties. A greater understanding of the potentialcontribution of research to policy and the constraints of policyformation would undoubtedly arise from closer collaborationbetween researchers and policymakers.

16

ReferencesAskew, I, Z. Matthews and R. Partridge (2002): Going Beyond Research: A Key IssuesPaper Raising Discussion Points Related to Dissemination, Utilisation and Impact ofReproductive and Sexual Health Research. Drawn from a two-day Meeting ofResearchers and Policymakers. http://www.socstats.soton.ac.uk/choices/workshop.

Bark, D.S. (1979): Overview of Research Utilisation for Population ProgramDevelopment from the Policy Analysis Perspective. A Strategy for Research Utilisation onPopulation and Family Planning. Korean Institute for Family Planning, Seoul, Korea.

Bernard, H.R (1994): Research Methods in Anthropology: Qualitative and QuantitativeApproaches. Second Edition, Chapter 10, 208-236. Sage Publications, Newbury Park,California.

ECDPM (2000): Using the Internet to share research and knowledge. InfoBrief 5,ECDPM Maastricht. http://www.ecdpm.org/pubs/infobrief5_gb.htm

Haaga, J.G and R.M Maru (1996):The Effect of Operations Research on ProgramChanges in Bangladesh. Studies in Family Planning, vol. 27 (2), p76-87.

Patton, M.Q (1990): Qualitative Evaluation and Research Methods. Second Edition. SagePublications, Newbury Park, California.

Porter, R.W and S. Prysor-Jones (1997): Making a Difference to Policies and Programmes:A Guide for Researchers. Support for Analysis and Research in Africa (SARA) Project.U.S Agency for International Development, Africa Bureau, Office of SustainableDevelopment.Washington, D.C.

RAND (1997): Population Matters: A RAND Program of Policy-relevant ResearchCommunication. RAND, Santa Monica, California, USA.

Smith, R (1993): Filling the Lacuna between Research and Practice: An interview withMichael Peckham. British Medical Journal, vol. 307, 1403-7.

Torres, F.D (1981): How do Research Findings get used in the Development Plan? Paperpresented at the 12th Summer Seminar in Population, Makati, Manila, June 28-July 3,1981.

Walt, G (1994): Health Policy: An introduction to Process and Power. Zed Books, London.

Walt, G (1994b): How far does Research Influence Policy? European Journal of PublicHealth, vol. 4, p233-235.

Weiss, C (1979):The Many Meanings of Research Utilisation. Public Administration Review,vol. 39, 426-31.

Worral, R (1972): Problems of Collection and Dissemination of Population Informationand Knowledge in Asia. Paper presented at the Economic Commission for Asia and theFar East, Second Asian Population Conference,Tokyo, November 1-13, 1972.

White, K (1993):We Couldn’t Have Asked for More: Lessons Learned in InformationDissemination.Technologies for Primary Health Care (PRITECH), Occasional OperationsPapers. Management Sciences for Health, Arlington,Virginia.

WHO (2000): Promoting Evidence-based Reproductive Health. Progress In ReproductiveHealth Research, No. 54, Part 1.

address etc

17