Document(s) 12 de 19

BACKGROUND TO THE DPRU—LEAN AND MEAN

From its inception in 1990, South Africa's Development Policy Research Unit (DPRU) had strong linkages with the anti-apart-heid mass democratic movement. In its initial phases, the central objective of the DPRU was to provide economic policy research to a movement that ostensibly was—and turned out to be—a future democratic government in waiting. Thus, from the beginning, the DPRU cultivated a strong relationship with policy-makers through its core activity, namely the production of academically credible policy research. This role became even more vital after 1994, with the advent of democracy.¹

Since that time, in the post-apartheid decade, relationships between networks linking the policy-making and the research communities were strengthened and new ones forged. While the key mission and objectives of the DPRU were elaborated before 1994, the research orientation and focus have shifted and kept abreast of the most pressing policy concerns of the post-apartheid and democratic period.

Taking the current political and economic environment into account, the mission of the DPRU is to:

• undertake high-quality, policy-relevant research
  
  
• maintain and develop effective networks with government, civil society and the research community in southern Africa
  
  
• engage in training and teaching activities
  
  
• participate directly in the process of formulating, implementing and evaluating policy.

In carrying out its mission, the DPRU has always sought to concentrate on developing and strengthening its fields of research. In the past, these research activities included labour markets and poverty, regulatory reform and regional integration. The latter two have gradually been subsumed by the fields of poverty, inequality and labour markets. This is partly symptomatic of the key challenge posed by labour markets and poverty for the viability of South Africa's democracy. In addition, however, it has been a deliberate strategic decision of the DPRU to pursue research excellence in a specific area of economic policy research, rather than spreading itself too thinly across a range of issues. This is critical in a country that is fairly well populated, with a number of highly focused policy research institutes. But it also means that DPRU's research activities have shifted from the unit's original research themes and gravitated towards research activities that can be broadly demarcated into six new themes. These themes, deliberately broad in scope, that capture the on-going research agenda of the DPRU are:

• education and skills development
  
  
• employment in the formal and informal sectors
  
  
• the economics of labour market regulation
  
  
• poverty and access to public services
  
  
• poverty and labour market outcomes
  
  
• economic growth, trade and poverty.

At the same time, the Development Policy Research Unit continues to adhere to its well-entrenched operational mandate, which constitutes its goals and objectives:

• foster high quality, policy relevant research within the DPRU
  
  
• train a new generation of research economists within the DPRU
  
  
• disseminate knowledge (research and its implications for policy) to decision-makers in government, the private sector and civil society.

The DPRU has only eight staff members. Five comprise its research component, with four working under the director who undertakes and supervises research production. The operations division has three staff members: the office manager; the secretary and the publications manager. The management team consists of the director and the office manager. Those at the DRPU recognize the virtue of its being, as popular speech puts it, "lean and mean". This philosophy governs the unit's principles of financial responsibility and the consciously embedded imperative of economic sustainability.

The DPRU's Modus Operandi

The DPRU, in pursuing its mandate of producing academically credible policy research, has as its client base numerous national and provincial government departments². These range from the National Treasury (typically the Ministry of Finance in a developing country) and the Presidency, to line ministries such as the Department of Labour, Department of Social Development and the Department of Trade and Industry. Given the solid track record of the institute, these departments often approach the DPRU directly to request an area of policy research be undertaken. However, in trying to ensure financial viability, the DPRU also—on a highly selective basis—competes for public tenders put out by the national and provincial departments.

In addition to delivering research products to our client base, the DPRU also pursues a program of capacity-building for public sector officials and members of the NGO network, union movement and civil society in general. This program is two-pronged. The first is to translate on-going policy-relevant, but technical research into more digestible material for a non-academic audience. For example, three times a year the DPRU runs an extension course called Labour economics and labour market policy, geared for a diverse audience composed of unionists, labour lawyers and public sector officials. The second aspect of the capacity-building program is to try, more formally, to build a sort of research–policy toolkit for technocrats within the bureaucracy. In this vein, we offer a training course called The analysis and measurement of poverty and inequality, which has as a key component the training of participants on a relevant statistical software package. This course is "harder" in that specific skills are taught, to sharpen the analytical tools at the disposal of the policy-maker.

While the first modus operandi of the DPRU is essentially knowledge production and the second capacity-building, the third of its operations is arguably the most important. That is the all-important task of disseminating and distilling knowledge. This is done through a variety of avenues, including regular workshops, an annual conference, a working paper series and a policy brief series. These are all governed by a regularly updated website. It is this particular function of the DPRU that provides numerous entry points for potential consumers of the unit's research output.

THE ROLE OF APPLIED POLICY RESEARCH INSTITUTIONS: THE NINE COMMANDMENTS

Given the unit's rich history and its location within one of the premier universities in South Africa, it has accumulated a significant quantum of knowledge and expertise in pursuing effective policy research. This can be expressed—for want of better labelling—as the "nine commandments" to which applied policy research institutes should adhere when trying to fulfill their mandates. While these truncated lessons apply directly to South Africa, there should be ideas here that resonate with similar institutes throughout the developing world.

1. Inform policy and policy-makers

One of the approaches to research, which distinguishes applied research institutions from consultancy firms, is that the former organizations undertake research principally to inform the policy process. This is distinct from undertaking research to influence the policy process. Hence, upon a policy issue being raised by a specific government agency, the task of the research group would then be to inter alia: sharpen the analytical parameters of the issue; provide exhaustive and value-added baseline information; close any obvious information gaps and question conventional wisdom on the issue. Using rigorous tools of economics, the research institution should attempt to provide a product that delivers information that complements and improves the quality and orientation of the decision-making process for the public sector institution.

For example, the DPRU has been engaged with the Department of Labour in a study on the incidence of atypical employment. The study was an attempt, through firm survey data, to both categorize and quantify the existence of atypical forms of employment in South Africa (see Lundall, Majeke and Poswell 2004). The Department of Labour is currently interested in seeing how it can legislate on this new development in the South African labour market. However, the information base from which the department was working was very poor. Hence the clear focus of the study was to improve on this information, with a view to sharpening the direction, and eventually the impact, of the intended policy interventions. Using local surveys, the study illustrated that there are numerous forms of atypical employment, ranging from part-time and shift work to out-sourcing, casual work and flexi-time workers. Such a nuanced definitional distinction around the particular policy issue—in this case atypical employment—had not even entered into the decision-making process of the relevant government department. So what the research managed to do was, in the first instance, suggest a direction for the policy intervention—namely ensure that the department would be focusing on what it viewed to be the most important subset of atypical forms of employment. Secondly, it improved on the quality of the policy-making process, in that the Department of Labour would have a much more nuanced assessment and appreciation of the incidence and typology of atypical employment, and indeed the particular factors that influence firms to engage in these forms of employment.

This is a very specific example, but it offers a general lesson. It may seem obvious, but it is still worth noting that the art of good policy research is really to inform rather than influence the policy process. In other words, policy research should complement policy decisions, and not be a reiteration or reformulation of these decisions. Once this is clear, the research institute needs to ensure that in the process of undertaking a piece of specific rigorous and credible research, it can help improve the quality of the policy decision and perhaps even the direction or nuance of the intended interventions.

2. Distil and disseminate knowledge

Research institutes, particularly those based at universities, often develop a poor reputation in policy circles for constantly delivering a high volume of unreadable or barely understandable reports to government ministries—the proverbial report that is read by no-one and simply occupies shelf-space. Herein lies the second lesson for policy institutes: the work only begins once the final report has been finished. The final report is the version that lends credibility to and provides details for the activities that will follow. In sum, these activities involve distilling and disseminating—in as deliberate a manner as possible—the results emanating from the research report.

Dissemination involves a well-known set of activities, including the publication of a working paper series, regular policy briefs and a functioning website with generous dollops of downloadable material. While these elements are well known, it is less clear whether all applied policy research institutes dedicate sufficient resources to this activity. The pursuance of the research contract, without ancillary activities designed to disseminate and share this information with a wider client base (donors, other government ministries, multilateral institutions), is in effect an implicit dilution of the potential impact of one's research. A key lesson then for policy institutes would be to dedicate specific resources on a continuous basis to ensure efficient and effective dissemination of their products. Implicitly, this is about a research institute having a well-functioning and well-resourced marketing division.

What is often less appreciated, however, is the need to distil information in the process of dissemination—the art of writing the one-pager, if you will. The policy community could legitimately argue that any product they receive that cannot be reduced to three or four key messages of substance is of no use to them. An example of how a final research report can be distilled comes from the DRPU, and a paper commissioned by the Presidency for the 10-year Cabinet Review Process commissioned a paper (Bhorat 2003b). The DPRU produced a 40-page report, then reduced that to just half a page by boiling down the key results that would inform the policy process. If these distilled ideas are powerful and relevant enough, they will permeate through the policy community. And if they do that, your research is already at the early stages of having an impact on policy. If you fail to deliver a distilled product, you can count on failing to have a policy impact.

It is these "headline" results that policy-makers incorporate into their decision-making. Ultimately then, any research institute that explicitly sees itself as policy-oriented needs to accept the notion that it is not just the ideas that are important in formulating policy; even more crucial is how effectively these ideas are transmitted to policy-makers.

3. Develop human capital

Research institutes are often ideally placed to recruit graduate students based within academic departments with which the institutes are associated. If appropriately managed and mentored, these students can often be trained, in the long-run, into highly competent research economists. However, for a significant number of new graduates, the research institute acts as a half-way station—the time they spend completing dissertations while working on a number of internal projects³. They will often leave the institutions after a stay of 12 to 18 months, either to pursue their own careers or perhaps when short-term research funding dries up. If possible, research units should try to help manage these young graduates' exits, perhaps helping find them appropriate positions in the public sector.

The public sector environment and its reputation as an employer may vary a good deal from one country to another, but in a number of developing nations there is a shortage of well-trained, experienced technocrats. Research institutions can provide these technocrats by acting as the link between the educational institution where these individuals are enrolled and the final employer. The DPRU has placed a large number of such individuals over the years, and indeed the public sector continues to request from the DPRU potential applicants for specific posts. This is clearly a "public good" that is being offered to government ministries, but it also builds the reputation of an institute—so that it is viewed as a competent enterprise providing a very good supply of young, well-trained technocrats. In addition, should this process continue over time, the institute widens its nodes of interaction with government officials, where a significant number of its ex-employees will be well-represented in ministries. This is the case for the DPRU, where members of its staff have over the years moved into senior positions in government, cementing good relations between researchers and policy-makers.

Indeed, this process needs to be carefully managed to ensure that the research institution does not lose staff members to the (often) better-paying public sector jobs, in cases where it does not want employees to leave. This requires careful management of more senior and experienced staff who should be retained with appropriate incentives, a flexible work environment and opportunities to advance their careers—a critical set of issues. Ultimately, what does distinguish an applied policy institute from the private consulting firm is that it serves as a partner (although not an uncritical one) on research with government ministries and other policy bodies. As such, this partnership should also extend, within reason, to providing value-added human capital to these bodies that very often lack skilled staff.

4. Build public sector capacity

It follows from the above that an additional focus of any applied policy research unit should be to build capacity within government. Government ministries often lack the requisite suite of skilled personnel to shape effective policy interventions. This is not the same as arguing that ministries should have research capabilities—as research is precisely what ministries should not be doing. The comparative advantage of government departments is in formulating, implementing and monitoring policy. But the experience (at least in the case of South Africa) suggests that ministries are either over-stretched to meet these obligations or simply possess staff without the right skills to fulfill their mandate. And it is here that research institutes can and should assist. This assistance should come primarily in the form of training workshops and extension courses. The former is represented by a hands-on, intensive (usually computer-based) course on a particular set of policy issues. These may include for example, a course such as the one mentioned earlier in this chapter, on analyzing and understanding poverty and inequality, or a course on trade and trade agreements. Policy-makers are often very receptive to such training. Evaluations of DPRU courses show that participants often leave feeling they have a clearer analytical framework than before, and equally important, a better set of tools with which to undertake policy work.

For example, many of the participants on the DPRU's Analysis and measurement of poverty and inequality course were members of the Department of Social Development, which is responsible for disbursing state transfers such as old-age pensions and child support grants. Many had not been previously exposed to the extensive literature looking at the welfare impact of these grants, nor indeed to the software that would allow them to track the progress of the grants, identify their recipients, calculate the poverty-alleviating impact of the grants, and so on. This course thus offered an essential capacity-building exercise for technocrats, transferring knowledge from the research institute to the public sector.

Extension courses are a second form of capacity-building. These involve a set of lectures to an audience of policy-makers, which rapidly offers them insights into on-going relevant research—both local and international—in a particular area. The DPRU's Labour economics and labour market policy course ⁴, for example, highlights some of the unit's on-going research into labour market issues. Topics include: the notion of "jobless growth" for South Africa; new trends in internal migration; quality of public education, and so on. Such extension courses for policy-makers benefit both audience and trainer. They introduce the research institute's program, market its product, to an important audience, while allowing policy-makers to rapidly absorb a condensed review of new local and international research material on a particular theme, building capacity in the public sector.

5. Be an honest broker to advance national policy debates

Academically credible research institutes occupy a key place in civil society; although they produce research for policy bodies, they do not—or at least should not—lose their credibility as an objective source of policy-relevant information. In societies with strong interest groups—and South Africa is of course no exception here—the presence of an objective, credible voice in the policy process is an invaluable one. That means that policy research units can carve out a particular niche as intermediaries within policy processes. The DPRU for example, has been involved in a number of key policy debates about the labour market, debates in which the union movement, business and government obviously may take differing positions. So the research unit plays a key role in trying to provide objective research to inform these debates among the social partners. Often these policy debates move forward largely because of research that clarifies the issue at hand and provides rigorous supplemental evidence. In the best-case scenario, policy research can actually break a stalemate in a policy debate.

The danger of course is that over time the institute's research products could actually help to cement in place a particular view on an issue. This may lead to perceptions that the institute is unfairly sympathetic to the views of one interest group over another, especially if the research in the past—albeit objectively—delivered verdicts perceived to favour a particular interest group. This may be very difficult to avoid, and an applied policy institution should be aware that its role as the purely objective interlocutor among interest groups might be difficult to maintain over time.

But this is not the only concern for research institutes. As they develop their public profiles and their presence in the market for research, donors and/or governments may begin—with varying degrees of persuasion—to use the institute to try and "front" or legitimize their policy positions. A ministry may want, for example, the institute to deliver a report on the benefits of privatization or the positive employment effects of trade liberalization, where the expressed intent may be to produce research that supports the policy position of the ministry. This seriously threatens the institute's role as the intermediary or the objective voice. It is critical that the institute manage these relationships carefully, keeping in mind that these actors (government and donors) do offer significant financial support to the institutes⁵.

6. Balance the benefits and costs of research networks

Research networks can operate both at the formal and informal level and each has its own challenges. Formal networks tend to be regional or sub-regional, and usually involve institutes doing a cross-section of policy research. Country networks are generally informal and develop through individual contacts and collaboration, without any strict form or cohesion.

Regional networks confront a number of difficulties. Member institutes may have no common thematic expertise, or have uneven levels and quality of expertise. They may be grouped because of their common region rather than shared research themes, and their work may also have varying degrees of policy relevance. All of these obstacles can make it very difficult to run and sustain an effective network. The absence of a champion or lead institute makes it more difficult to sustain a viable network.

Research units are also often organized and run on tight timelines, with deadlines for delivery and short-term funding. In such an environment, membership in a network can sap time, energy and human resources and become a burden for individual research institutes. On the other hand, networks do offer a rare opportunity to engage with peers and institutions outside one's own country and to learn from their experiences. They can offer useful cross-pollination of ideas in diverse areas, including: operational, financial and human resource management; dissemination activities; fund-raising protocols; and proposal writing. Perhaps the key role for an institution in a network—if it is a strong, well-functioning one—is to assist weaker partners in the network in a way that does not draw too heavily on its own internal time and resources. Ultimately though, networks are an important part of the research environment and institutes should try and involve themselves in these activities, taking care not to sacrifice their own internal activities.

A local network usually operates far more loosely and tends to provide a higher rate of return to its member institutions. They offer the obvious advantage of being able to spread information evenly among members, and without great cost. An institute may become aware of a donor's new area of research focus or a government ministry's pending call for tenders through an informal network that permits constant interaction with other domestic policy units. Such networks are also useful for institutes with different areas of expertise, allowing them to come together to collaborate on projects to which they can all contribute, and perhaps qualify for funding of a project that individually they would not obtain. In a slightly different vein, being part of an informal network offers the opportunity to outsource segments of a project to external qualified individuals or institutes. This offers significant long-term advantages, saving the institute costly up-front commitment to human resources they cannot afford, or which cannot be sustained if there is not a constant stream of projects or funding is not guaranteed, and it rarely is in the current funding environment. Finally, members of such informal networks can help control quality of each other's research. This may happen more formally, when, for example, the client requests a formal evaluation of a product. But as an internal member of the policy research community, units can and should ensure that the quality of work that is delivered to clients does justice to policy formulation and to the policy process in general.

7. Paint the "big picture" for policy-makers

Policy-makers need baseline information but they also require data that give them the "big picture". This means they need standard sets of indicators on a variety of macroeconomic, trade, labour market, financial and other themes. In some countries in Africa, the national statistical agency may not function well enough to distribute these data. Clearly, the research institute is ideally positioned to serve as the locus—for a national and international audience—for this type of data. Such "big picture" data can become a springboard for future, more intensive research and so can be viewed as a sort of market-entry strategy for a new institute. In countries where the statistical agency does function well, research institutes may also need to produce more value-added indicator data. For example, in the South African context, while Statistics South Africa does publish employment and unemployment data regularly, these data may not always fully serve the needs of policy bodies. So the DPRU often provides data on employment distributions by sector and occupation—data that can be garnered only through manipulation of the unit's records. SSA seldom publishes this type of data. These kinds of standard indicators can also include more obvious statistics that policy-makers often look for, such as the headcount index or the Gini coefficient, which are seldom found in a statistical agency's releases.

This broad-based and extensive indicator information should also be disseminated appropriately, preferably through an institute's web site where it is easily and freely accessed from anywhere and by anyone. Such information can be invaluable in attracting donors to an institute; newcomers in a country who want a quick overview will visit the web pages of a relevant institute, if it is appropriately marketed. These generalized "Economy by Numbers" data that draw the big picture as well as small ones, can provide newcomers to a country, foreign donors and government ministries with vital but straightforward information, thereby raising the profile and relevance of the research institute.

8. Collaborate wisely with international researchers

Research institutes are ideally placed to engage constantly with international researchers. This can be with researchers working in their particular field of interest, who may not have special knowledge of or interest in the country in which the institute is based but who have made significant contributions to the body of economic knowledge. Interaction can also be with international researchers who do have a special interest in the country, often individuals who have built their careers as renowned experts on a particular country or region.

Accessing and engaging with these researchers is a key role for applied policy research institutes. This can be done in two main ways. One is to attempt to fund projects that would formally draw these individuals into the given project as collaborators. Second is to provide these foreign experts with a platform for engaging with local government ministries and other senior members of civil society. For the former, the institute needs to ensure that it is getting an external individual who adds value to the project, and is not simply a substitute for a researcher who could be locally hired. That is, they should bring with them particular skills that are hard to find locally. The DPRU, for example, is working on a project on schooling that will engage a number of externally recruited researchers. They bring with them a very high level of experience, knowledge and specific expertise that would be difficult, if not impossible to find locally. An additional component of this collaboration would be to ensure that the international recruit transfers some skills to the institute. This skill transfer is always stipulated formally in contractual agreements, but such transfer rarely takes place. For this reason, the institute should organize formal sessions during which particular skills are transferred, and ensure that publications that emerge from the collaboration are jointly authored. There are cases when an international researcher may not offer a unique set of skills or perspective, and research institutes need to avoid this type of collaboration.

Apart from collaboration with international researchers within the institute, it is also important to facilitate their engagement with local policy-makers. This can take two forms. The first is to arrange training workshops for policy-makers at which international researchers are some of the key presenters. They bring with them a fresh perspective and new ideas for local audiences, and this sparks useful interaction. A note of caution here: the international researchers must be informed of the nature of the audience to ensure that the training is well-suited to their needs and interests.

The institute should also arrange a series of high-level dialogue sessions bringing together their foreign experts and local policy-makers who stand to benefit from their particular expertise. For example, an institute could invite an internationally renowned macroeconomist to have a dialogue with Treasury and Central Bank officials. This obviously builds the profile of the institute, but more importantly, it facilitates potentially fruitful engagement between a highly-skilled foreigner and local policy-makers.

9. Take advantage of useful new methodologies and approaches

An advantage to a research institute that is based at a university is that the institute has ready access to new techniques and advances in the economics literature, although this can also be the case for research units off-campus. This access comes in three forms. First is that universities tend to subscribe to relevant and recent international and local scientific journals. Second, universities permit engagement of researchers and academic staff. And third, research institute staff can lecture in graduate courses; individuals teaching graduate courses engage with the academic material in a way that they may not when they are conducting a research project. This builds technical capacity for researchers at all levels. For example, teaching a course on survey econometrics will build the capacity of the researcher, and introduce him or her to new methodologies in a far more efficient and effective manner than any given research project is able to do.

The secondary effect of keeping abreast of new methodologies in this way is a solid return to the research unit—and to the policy process. For example, the use of what were to the South African research community relatively new techniques in panel data estimation played a key role in informing policy-makers on income and job mobility in post-apartheid South Africa. Another example of the benefits of new techniques in South Africa are the approaches being developed to understand intra-household behaviour and activities. These new approaches will be instrumental in understanding the impact of government's social grants system, particularly the child support grant. Recent econometric advances—the mixed logit models that are in use in developed countries—will no doubt add more nuance to research in developing countries.

However, there is an art to adopting new methodologies. Namely, they should only be adopted if they offer two essential advantages: they add new or more nuanced policy information to the analysis; and they deliver a significantly more robust analysis, even if their contribution to policy is only marginal. Where the gains from new techniques are not substantial in either of these two ways, then their adoption should be approached with circumspection.

CONCLUSION

Obviously no listing of the key components of an effective policy research unit can be exhaustive. As stated at the outset, this overview is offered as a case study from South Africa and clearly, the emphasis on each of the prescribed roles for an applied policy research institute will vary for other countries, regions, research units with different areas of interest, and indeed the capacity and financial well-being of the institution. However, it is likely that for a medium-sized institute undertaking economic policy research in a developing country, these "nine commandments" do apply to some extent, and would contribute to an institute's strategy to deliver quality policy research in an efficient and effective manner.

Notes

¹ For a more detailed and broader overview of the role of research in policy-making in late- and post-apartheid South Africa, see Bhorat 2003a.

² South Africa has nine provinces, each with their own provincial line ministries. In theory, the national departments set out the broad frameworks for specific interventions, while the provincial departments' brief is to implement the policies within this broad nationally-crafted remit.

³ Ideal, of course, is to have the graduate student work on a project that eventually can be reformulated with minimal effort into a dissertation.

⁴ Details for this labour markets course can be found at http://www.commerce.uct.ac.za/dpru/MESP_Course/EXTENT2004.HTM; and for the poverty and inequality course, at http://www.commerce.uct.ac.za/DPRU/TechnicalAssistanceCourse2004.htm

⁵ For an excellent insight into some of these difficulties within a South African context, see Cassim (2004).

References

Bhorat H. 2003a. South Africa: policy-orientated research on labour markets and poverty, IN: Phillips LC and Seck Diery (eds). Fixing African economies: policy research for development. Colorado: Lynne Rienner Publishers.

______. 2003b. Employment and unemployment trends in the post-apartheid South African labour market. Paper commissioned by the Presidency for the 10-year Cabinet Review process. Cape Town: Development Policy Research Unit, University of Cape Town

Cassim R. 2004. Reconciling the mismatch between the producers and consumers of economic policy research: the case of trade and industrial policy strategies (TIPS) in South Africa. Johannesburg: Trade and Industrial Policy Secretariat. Unpublished mimeo.

Lundall P, Majeke A and Poswell L. 2004. Atypical employment and labour market flexibility: experiences of manufacturing firms in the Greater Cape Town Metropolitan Area. Report to Department of Labour, South Africa. Development Policy Research Unit, University of Cape Town.

Document(s) 12 de 19