by Erik W. Thulstrup Education and Social Policy Department, The World Bank ABSTRACT

It seems increasingly clear that scientific research and research training is essential for the development process. A major difference between developing and developed countries is related to their ability to select, create, modify, adapt, or use advanced technologies efficiently. The high complexity and frequent changes in the most efficient technologies have increased the quantitative and qualitative need for research trained manpower in many developing countries, but these countries are often unable to carry out high quality research training on their own. At the same time, overseas training is becoming increasingly expensive and is in some cases less relevant to domestic needs. At least a partial solution to these problems can be obtained if relevant research activities in developing countries are improved and expanded.

The present paper summarizes some of the work which was carried out in the S&T Work Program (part of PHREE and later ESP) in the World Bank during 1992-3. After discussing the triple research needs of any country: basic scientific research, applied research, and mechanisms to improve the utilization of research output, a survey is given of research financing, research communication, individual research incentives, and research policy. A summary is provided of Bank experience and of pitfalls in Bank projects supporting research capacity building, especially problems related to lack of policy actions and output monitoring. Finally, a list of major problems in scientific research capacity building in developing countries is given, together with possible solutions.


Developing countries, with 80% of the World’s population, account for only 2% of the global expenditures on scientific research and for an even smaller share of the research output, which is the quantity of direct importance. Leading industrialized countries spend close to 3% of their GNP on research and development, while most developing countries, with some major exceptions, spend well below 1% of their relatively much smaller GNPs on research. The problem is further increased by the fact that much of the needed hardware and consumables for scientific research (instruments, materials, chemicals, communication) also in absolute terms tend to be much more expensive in developing than in industrialized countries.

The reason for the low priority given to research is not that developing countries do not need any significant scientific research capacity, but rather that the returns of investments in research often are indirect and that they are visible only several years after the research capacity has been established. In those cases where reliable rates of return for scientific research can be produced (primarily in agriculture) they tend to be high, often 30-50% (Mansfield, 1993).

The most efficient technologies, both in agriculture, manufacturing, health, transportation, communication, and environmental fields, are changing fast, and this pace of change is expected to continue. Using last decade’s, sometimes even last year’s, technology will often not satisfy the market demands. In recent years, many developing countries have learned that not only local development of competitive technology, but also successful import of new technology is very difficult. The latter requires both careful selection and considerable ability to modify the imported technology locally. At the same time, the internationalization of manufacturing, trade, and investments has helped make new, efficient technologies globally available shortly after they appear on the market. This has had a considerable impact on the global competitive climate.

Today, highly efficient technologies, even low cost technologies and technologies adapted for local use, tend to contain a large amount of research based knowledge. This makes not only the development of competitive technology, but also technology import and modification very demanding. In most cases, local human capability at a high level is required, both with respect to knowledge of the field, which the technology deals with, but also in terms of analytical and decision making skills. Such capacity will result from research training activities within the field in question. A increasingly important mechanism for knowledge transfer is based on graduate students, who are involved in academic research. After graduation such students often become major carriers of new knowledge from the research laboratory to industry. According to Mansfield (1993), university laboratories that received industrial support usually provided the industries with both research results and this kind of specialist manpower; in the United States, the percentage of research groups from which graduates took employment in the supporting industries was high, from 64% in the field of chemicals to 87% in information processing and 91% in petroleum.

The globalization of production methods and the need to use up-to-date technologies in order to stay competitive have therefore made the availability of research trained manpower a key issue, both in industrialized and developing countries. Since efficient and high quality research training requires efficient and high quality research activities, it has become necessary for all countries, including the poorest developing countries, to further define and strengthen their policies with respect to research and research training. Availability of highly trained science and technology manpower is becoming one of the most important criteria when decisions are made in the international market about the placement of high value-added production facilities. At the same time, the technological development has decreased the competitive advantage of low-cost labor, which for many years was the economic foundation of attempts by developing countries to attract manufacturing activities. For many years, Hong Kong had considerable success using this strategy, more recently relying extensively on labor in China. Realizing the new demands, it has recently decided to expand high level science and technology education and research very fast and has invested heavily, both in the new Hong Kong University for Science and Technology and in expansion of the existing universities (Wu, 1992). Other countries in East Asia, especially Korea and Taiwan followed a similar strategy more than 20 years ago, and were in many ways very successful (Wu, 1993).

Realizing that “development in our time is inseparable from scientific advance” (Hoffmann, 1992), developing countries may try to learn from such examples. They might benefit particularly from the more recent experiences in Hong Kong. However, each country will have to identify its own comparative advantages and define its own strategy (Thulstrup, 1992). On balance, due to their extensive human resources, many developing countries still have good opportunities to attract higher value-added activities, but they must realize that these depend increasingly on the ability of the countries to produce highly skilled, including research trained, scientific manpower.


If it is assumed that the most important product of research in developing countries is research trained manpower rather than research results, then those research activities that help train new talent become of the highest importance. Many developing countries have invested large amounts in expensive research institutions outside universities. Often such institutions are not very productive with respect to research results, but even those which are research productive tend to provide little training. This is the case in spite of the fact that they often have expensive facilities for high quality research which universities frequently lack. Unless special conditions exist, it will generally be advisable to concentrate research investments where training takes place or to move research training to places where both research facilities and activities are strong. There has often been a tendency among donors and international organizations working with developing countries to favor investments in selected fields of applied research, rather than in research training and basic scientific research. The reason is usually a wish to produce visible development results fast. Often applied fields offer special opportunities, e.g. development of new consumer goods based on the natural products available in a country or other of its competitive advantages. However, the sustainability of applied research and development activities will often suffer unless it is supported by a flow of new, research trained talent and by a national research base in the science fields which have made the applications possible. Therefore, selected development of those applied fields which support key economic activities in the country or which makes it possible to fully benefit from national competitive advantages, must be encouraged. In order to support research and development in applied fields, the basic sciences on which they are founded must also be strengthened. Many new, efficient technologies are very directly supported by recent progress in the basic natural sciences (new materials, some biotechnologies, laser based technology, etc.). This also means that these basic sciences function as an important supply source for research trained manpower, which is needed to carry out applied research and development within agriculture, health, manufacturing, communication, the environment, etc. The basic sciences also supply much of the knowledge and basic training provided to students in applied fields: For example, more than half of the courses given to chemical engineering students during their undergraduate training are likely to be in basic chemistry, biology, physics, and mathematics. A third necessary condition for enjoying the economic benefits of research activities is the insight, managerial capability, and incentives needed to apply existing, research based knowledge in the development process. Application, even of potentially very valuable knowledge, does not come by itself. Often governments and manufacturing industries are hesitant to ask researchers what they can provide, either because they do not expect any useful answer or because they feel that they may not like the answer (Box 1).



During a visit by a Bank mission in 1991 to a large governmental applied research institution in Poland, the Director complained that new staff with a solid research training had an unfortunate urge to change both products and production methods. This attitude was considered a nuisance by the Director, and the Institute had hired very few researchers with Ph.D.s (only around 3% of the staff had Ph.D.s), in spite of the fact that Polish universities and institutes under the Academy of Sciences produced many excellent, young graduates from relevant Ph.D. programs.

Most of the industries supported by the Research Institute were already at the time in considerable difficulties because of insufficient product development and inefficient production methods, but these problems had not been visible earlier because of the lack of competition in the market.



There have been many attempts to change this situation. In recent years, university-industry cooperation has become increasingly fashionable, also in several developing countries. An international survey of mechanisms for university-industry research cooperation has been compiled by L. Parker (1992), who concludes that:

(i) It is demanding. Universities must be able to perform research of good quality and industrial relevance, industries must be willing and able to use the results constructively, and usually also governments must take part, playing a very difficult, constructive role, if the cooperation is to succeed;

(ii) university-industry cooperation generally seems to be based on the specific academic culture and values which are associated with Western universities; these models may not easily be transplanted to other academic cultures, at least not without considerable modifications; and

(iii) the direct outcomes of university-industry cooperation, e.g. in the form of patents, may not be the most important. In many cases, the information on industrial needs and conditions in the labor market, which is acquired by university staff and students in connection with the cooperation, may be much more valuable. Students involved in the cooperative work are likely to be better motivated and trained for work in industry, and industry may be able to spot needed talents among the students at an early stage.

University-industry cooperation rarely develops fast into successes. It often takes time for both partners to learn how to cooperate constructively. “Learning by doing” seems to work best (A. Kornhauser, 1992).

Thus, three ingredients are needed in order for a country to benefit from the opportunities of research based knowledge: (i) Broad basic science research and research training capacity, (ii) selected applied research and development capability in fields of particular importance to the country, and (iii) methods and incentives for practical use of the knowledge in the development process (Thulstrup, 1992). For an example of the importance of the latter ingredient, see Box 2.



Basic and some applied research has reached a high level at the best universities and some governmental research institutes in Poland. Polish researchers have made important contributions to the global knowledge in several fields. However, this capability was often not transferred to the manufacturing sector. The high quality research and research training did little to improve the economic conditions in Poland, even at a time and in fields where research based knowledge had become extremely valuable. For example, although Poland produced close to 2% of the World’s scientific papers in mainstream journals in the field of polymer science, and received well over 1% of the citations (a measure of how useful the research is to other researchers) in this important field, it only held 0.06% of the patents within rubber and plastics products. In these fields, it ranked 10th in the World in research papers, 12th in citations, but only 21-26th in patents. Similarly, Poland ranked 6th in the World in chemical engineering research papers, 11th in citations, but ranked only around 27th in patents in fields like biotechnology, pharmacy, and petroleum (The World Bank, 1993).




While typically 35-45% of all research and development expenditures in industrialized countries take place in the public sector, a much higher share of the research expenditures in developing countries tend to be covered by state agencies, especially public universities. This is a reflection both of the fact that industry in most developing countries considers itself less dependent on research based knowledge (which may not always be true) and often is financially weak, but also of the recognition in the public sector of the need to secure a minimum level of research based training capacity. The latter need is surprisingly widely recognized, also by countries in the Third World; globally, the lower total research expenditures are, the higher a fraction tend to be spent at universities (Thulstrup, 1992).

Often public sector support for research in the Third World is spread out: Each researcher is given a small amount, usually a higher amount in research institutes than in universities. In such systems, good researchers will often have insufficient resources to be productive and no incentives for active research result from the financing, as discussed later. Even in other systems, where financing of research is based on performance, the training impact of research activities are sometimes overlooked -research grants do not contain support for research students and may often be given to activities in which no significant human resource development takes place.

Funding for research in developing countries is frequently provided with little regard for the expected outcome of the research; most of the emphasis is on the input. In World Bank projects, this attitude is sometimes reenforced by the Bank’s heavy involvement and intensive monitoring during the project period, when investments (inputs) take place, compared with the lack of Bank monitoring after the project period, when outputs could and should be expected (see later).

In most industrialized countries, tuition fees from students pursuing research degrees in scientific fields cover only a very small share, if any, of the often high costs of their training. If such fees exist, they are often covered by the students’ advisors through their research funding. This is a consequence of the important role graduate students play for the research productivity. In some cases the need to pay high tuition fees has had an unfortunate effect: It discourages researchers from accepting graduate students – post-doctoral fellows may be less expensive and are even more productive. The wish to get access to outstanding students has in S&T fields frequently led to a situation where tuition fees are covered by other sources and where students are paid for providing services related to their studies such as laboratory work or undergraduate teaching in their area of specialization.

For developing countries, this may lead to two conclusions for domestic and overseas graduate students in science fields: (i) The financial contribution of domestic graduate students in science based fields is likely to remain a relatively minor source of revenue, and (ii) highly qualified developing country students overseas might often be able to raise at least part of the funds needed for their training within the foreign university.

Many developing countries have invested heavily in such training of researchers abroad. In most cases, the developing country students attend graduate schools in industrialized countries, rather than in other developing countries. The developing countries often pay the full costs, for example through World Bank loans, covering both subsistence and tuition, without even attempting to find support as mentioned above. Therefore, the total costs for overseas degree studies for developing country students tend to be very high, typically around $25,000 a year (and well over $100,000 per Ph.D.). In comparison, industrialized country graduate students in scientific fields rarely pay tuition fees in other countries; a recent survey shows, for example, that among Danish graduate students in the US only 11% pay tuition fees (The Danish Research Academy, 1993).

Although overseas studies may be necessary to build up local research capacity in key fields in a developing country, they frequently have severe drawbacks. The traditional overseas programs, for example under World Bank loans, often provide more support for universities in industrialized countries than for the domestic, developing country universities, which frequently benefit little during the period of overseas studies, since they are not much involved in the process, even when it is their own staff that is being trained. The graduate research projects often do not deal with the subjects of highest priority in the developing country, and in some cases the returning young researchers are unable to use the training they have received, because of lack of relevance, facilities, or local ntellectual stimulation in their field of specialization. Finally, long term overseas training without much contact to the home base may easily lead to brain drain of young researchers, which for some countries is a severe problem. In recent years, several bilateral agencies have realized this and have started emphasizing the involvement of domestic institutions in overseas research training programs, e.g. by supporting institutional North-South twinning, often with research training carried out jointly by the institutions involved (Danida, 1992; Gaillard & Thulstrup, 1993).

In many countries, the government supports large amounts of research and development in private industry – in the USA, for example, the Federal Government alone covers around one third of all industrial research and development expenditures; this kind of governmental support also takes place in some developing countries. On the other hand, industry rarely provides a large share of university research expenditures – 7% in the USA, 2% in Japan, and almost nothing in most developing countries. However, support from industry and other outside agencies (especially in Africa, bilateral support of research accounts for much of the investments) may be significant in specific scientific fields. In addition, industry often provides personal incentives to selected researchers.

In practice, the personal rewards from industry to individual researchers may be either conducive for good research – e.g. by providing incentives for further development and applications of advanced research results, or destructive – by rewarding researchers financially for carrying out routine tasks, thereby distracting them from their research and teaching duties, in particular from providing sufficient support to their research students. In developing countries, bilateral and international organizations occasionally create problems of the latter kind by making consultancies excessively attractive to a frequently very small group of qualified university researchers in the key fields of interest for these organizations. This may severely limit the training of new experts in the fields in question.

The effect of occasional outside support for individual researchers is particularly strong in developing countries because the regular salaries of most researchers tend to be insufficient for sustaining an acceptable standard of living. Having second incomes is often a necessity. Only rarely are the second jobs in developing countries conducive to the research activities; usually they will take time away from research and (graduate) education. Lack of time and incentives for research may be the most serious problem for many developing country research communities. Surprisingly often, large research investments are made in developing countries, especially for buildings and other hardware, without making sure that the researchers involved will be able to give up such other tasks and use the new facilities full time.


Scientific knowledge increases today to a large extent through global cooperation and competition. Contributions of individual researchers and research groups (outside certain industrial and defense related fields) are exchanged daily at the international marketplace through electronic mail, fax, scientific meetings, and research journals. The latter have traditionally been the core of scientific information exchange. It may seem like this kind of information is available at relatively very low cost, although journal subscriptions often place a hard burden on developing country researchers. However, in reality it is difficult for users of this kind of information to benefit fully unless they, e.g. the developing country researchers, are active and productive members of the international community in the given field. For the individual user, writing science is as important a part of the process as reading it. Researchers who do not present their own findings to the international community will not receive the important quality control provided through journals or directly by their peers and will remain outside the global, informal networks which exist in most fields. This imposes severe limitations on their opportunities for quality improvement. The increasing importance of electronic communication has often made matters worse for developing country researchers, since the required facilities are often not available to them. In the long term, however, electronic communication may become a very valuable tool for researchers in developing countries, since it can compensate their traditional handicap of being far away from other researchers in the field.

The situation for many developing country researchers is that they, in spite of considerable talent and often also strong backgrounds from research training in good laboratories, publish little in mainstream research journals, do not use electronic communication, and to a large extent remain outside the international networks in their field. The situation is illustrated by the production of mainstream journal papers in African countries (Table 1, Eisemon & Davis, 1993). The largest producer in Africa, Nigeria, produces only a fraction of what is produced in small, highly developed European countries and the development in Nigerian research as in many other African countries has in recent years been negative. For example, 1200 authors from Nigeria produced papers in international mainstream research journals during 1985 compared with only 800 during 1991.

In the home country, developing country researchers often find a large number of local research journals which appear with irregular intervals, have less than ideal review systems and are read by few researchers in the field (Thulstrup, 1992). The circulation of such journals among researchers in the fields is often poor, even within the country. Nor do most of such local journals act as an efficient communication channel to local industry. In many developing countries, merging such journals into a much smaller number of national or regional journals might be of immense benefit to local researchers and might at the same time save money. During the 1970s, the Nordic countries went through such a process, transforming local research journals into regional (in reality, international) research journals; the research communities that were affected by this transformation have now developed into some of the most efficient in the World.


It has often been assumed that the largest obstacle to active research in developing countries is lack of modern equipment. Also the small amount of time many developing country researchers can invest in their research is generally acknowledged as a major problem. However, in cases where both equipment and time were provided, research often remained lackluster. One important reason might be that individual incentives for active research were missing. It has often been noticed that researchers generally respond more effectively to incentives than to orders.

The most important personal incentives for researchers may be their wish to gain fame and honor. Intellectual rewards of many kinds are also very important, for example those resulting from participation in strong team research, which today often provide efficient incentives. However, financial incentives are of a comparable effect and are often simple to provide as part of an efficient research policy. But this is frequently not done in developing countries. Often promotion, salaries, travel funds, research funding, and other possible sources for incentives to researchers in developing countries are provided with only limited regard to the amount and quality of the research work they perform. As mentioned earlier, this has two consequences: (i) The usually very modest funds available are spread out to all researchers so that none are given a critical mass of support and (ii) the researchers are not provided with the kind of incentives for efficient research which would result from a funding system where rewards were based on earlier accomplishments.



Total No. of Per Million

Country Authors/Papers* Population

1981-90 1981-90

Cameroon 606 58

Ethiopia 1,012 23

Ghana 663 50

Kenya 4,109 194

Nigeria 11,778 114

Sierra Leone 175 46

Sudan 1,481 66

Tanzania 1,127 49

Uganda 375 25

Zimbabwe 1,597 184

Denmark 39,300 7,700

Sweden 74,800 8,900


Sources: For Africa: Eisemon and Davis, 1993. The numbers are sums of annual numbers of different authors, listed by Eisemon and Davis (1993). These may be expected to roughly represent the number of papers published. For Europe: Thulstrup, 1992. Note that the original database used, Science Citation Index, has a language bias which favors English speaking countries.

Thus, in addition to sufficient salaries to researchers, acceptable physical facilities, funds for operation and maintenance of these, support for communication with other researchers (publications, electronic mail, fax, conference and other travel), incentives must be in place which secure that individual researchers feel that their efforts are noticed and rewarded. This is often the case if the amount of support to each research project reflects the quality and efficiency of the research and dissemination activities.


Donor dominance of research activities occurs in several developing countries, especially in Africa. Often donor support is not coordinated with the national policy, if such a policy exists, and when donors dominate, it becomes difficult for any government to pursue a consistent research policy. Most donors prefer to contribute in highly visible, directly development relevant fields, and want results within a few years. These preferences are not well suited for research capacity building, which takes many years and includes establishment of a considerable research infrastructure which is not highly visible. For example, the national capabilities in mathematics and statistics as well as basic natural sciences is essential for many important applied fields (agriculture, engineering, environment, health, etc.), but are still by many not considered clearly significant for the development process.

For these and several other reasons, it will generally be necessary that developing countries develop research and science policies of their own. Such research policies must have a long term perspective; shortcuts are not likely to work. Since research needs and capacities in any country are interrelated, it must also be comprehensive: It should take into account research activities under all ministries, all local authorities, and in all sectors, including the private. In particular, the education and training activities, which supply other research and development activities with the needed manpower, must be carefully coordinated in order to provide the needed outputs without excessive waste.

An important part of the implementation of any research and science policy is monitoring and evaluation of research related activities. Evaluation procedures must be established at the local level (in connection with hiring and promotion of researchers, distribution of local research funds, etc.) and at the national level (competitive distribution of research funds through research councils, large scale evaluation of major fields as input for science policy decisions). Furthermore, convenient and informative science and technology indicators must be defined and carefully monitored. In this connection output indicators are particularly important (research papers, citations, patents, number of Ph.D.s produced, etc.). For a case study of developing country research based on bibliometric indicators, see Gaillard (1992).

Monitoring science indicators and evaluating research activities are today common practice in several industrialized countries; these activities have had considerable success in the Nordic countries. A parallel activity within the field of higher education indicators is emerging (Kells, 1992) but is not quite as established yet. International science and technology indicator work has been carried out for several years within the OECD cooperation (Kruytbosch, 1993), but science and technology, research, and higher education indicator activities are much more rare in developing countries.

Examples of regional science and technology indicator projects in the Pacific, Asia, and Latin America are given by Kruytbosch (1993). One weakness in most indicator systems, including much OECD work, is that they concentrate on inputs and often partially or completely neglect output monitoring. UNESCO has for many years attempted a collection of Science and Technology Indicators in the Statistical Yearbook, but the information is in many cases outdated and unreliable. In an effort to improve the data collection in member countries, UNESCO hired in 1992 two consultants who made week-long visits to individual countries in order to work with local staff responsible for the science and technology data collection. However, unless such efforts are ncreased drastically, it may take decades before reliable indicators become available in most developing countries.


The World Bank has provided support for research activities in developing countries, especially through agricultural, industrial, infrastructure, and educational projects. In the latter, most support was given to research within science based fields. Support for science and technology in higher education projects has increased dramatically since 1980, especially in Asia, where the total project costs grew from $131 million in the 1970s to $4,525 million ($1,334 million in loans) in the 1980s (Muskin, 1992). All other regions, especially Africa, received much less. Globally, World Bank project costs for science and technology in higher education now are expected to amount to between $500 million and $1 billion annually, and the regional distribution, at least outside Africa, is expected to become more even.

Among the 32 higher education science and technology projects that were initiated between 1970 and 1991, about one third had explicitly stated research and development objectives. Such components occurred in all regions except Africa, where none of the four projects contained research objectives. The greatest expenditures in the projects were for buildings and equipment, which on the average accounted for around one third each of the total. Among other budget items, graduate training, especially overseas, was the largest. Also, operation and maintenance were given considerable resources.

Most projects aimed at direct support for industry had scientific research components (Muskin, 1992). Among the 13 projects supporting science in industry, which were initiated between 1970 and 1991, essentially all had explicitly stated research objectives; 12 supported research institutions and 5 also universities. None of the projects occurred in Africa. The total costs for the projects were over $2 billion, about half of this amount came from World Bank loans. Also lending for science in industry is expected to increase; the projected annual costs for industry science and technology projects during the 1990s are around $400 million. Compared with the higher education projects, much less was spent on buildings (7%), while a similar share (one third) was spent for equipment. An activity not included in any of the university projects, on-lending, here in support of project activities in individual industries, was given 26% of the funds.

While some research support in Bank projects has been given successfully and has helped strengthen good research practices in borrower countries, there have also been failures. Among the main reasons for the failures are: Neglect of output goals and a lack of balance between project expenditures and policy actions (Thulstrup, 1992a):

Lack of output monitoring. During the preparation of a World Bank project, all efforts tend to be concentrated on inputs, e.g. buildings, equipment, scholarships, travel, etc. Output goals – research results, training, etc., and ultimately building of research capacity – are often only mentioned in the project document, but not used initially as indicators of project success. After the project start, monitoring concentrates on inputs and adherence to a time schedule for expenditures is considered an important indicator of success. Often output monitoring systems are not put in place in time for the first measurable outputs to appear and definition of long-term output indicators are sometimes completely neglected. Finally, after the loan amount has been dispersed, Bank involvement is considerably reduced, but the input based monitoring systems used by the Bank are often adopted by the borrower country, while output monitoring is never introduced.

Lack of policy actions. Since the main parameter, carefully monitored in World Bank projects, is the loan amount, there is a tendency to concentrate on activities that cost money and leave other, equally important activities (especially on research policy matters) to others. Since a strengthening of research capability in a country in most cases also requires a strengthening of the national research policy to be effective, this strategy is often shortsighted.

Project failures may show up in many different ways, but they are often only clearly visible if output indicators are defined and monitored as discussed above. Among the most common failures in World Bank research capacity building projects are:

(i) Provisions of research equipment have frequently not worked well. Often the equipment has not been matched to the local conditions, needs, know-how, and services available. Maintenance and spare parts have often caused problems. Many kinds of commercial research equipment is under almost continuous updating; getting last year’s model may not allow the researchers to compete internationally. In many projects, the researchers who will be using the equipment have had little influence on the actual choice (due to the competitive bidding format) and have had no opportunity to negotiate price and technical specifications with the producer. Local suppliers are sometimes unable to provide the service promised in the contract and do rarely have a sufficient stock of spare parts to prevent extended periods of breakdown. Delays in the procurement process have sometimes resulted in delivery of outdated equipment. Finally, the price of both equipment and spare parts tend to be much higher than in industrialized countries,

(ii) sometimes funds for operation and maintenance have been insufficient. Although the need vary considerably, one might count on an average annual expenditure of at least 10% of the cost of the equipment for reasonably active operation and maintenance. Also, much equipment requires frequent replacement of certain parts with a limited lifetime. Often such parts for a few years’ use have not been supplied together with new equipment, primarily because experts in the use of the equipment have been too far removed from the final purchase process,

(iii) often projects have not included provisions for research time for the developing country researchers, who have therefore not been able to concentrate sufficiently on their research work, as discussed earlier. In addition, incentives may be lacking – both World Bank projects and the national research policies have frequently been designed so that it makes little difference to the individual researcher whether or not the research is successful and produces outputs like well disseminated quality results, research training, and support for local users, and

(iv) some Bank projects have made the assumption that even expensive overseas research training in industrialized countries is more cost-effective than domestic training. This has sometimes been based on an analysis that covers a too short time span. Since the long- term need for research training is huge in most larger developing countries, such countries may benefit from an early build-up of graduate education, both for economic reasons and in order to ensure sufficient relevance. Regional solutions to the need for research training have occasionally been overlooked; they may be worthwhile for any country. Even within the European Community, which include several research superpowers, large efforts are made to strengthen the regional cooperation within research training.


In summary, developing country researchers, research managers, and research policy makers face a large number of severe problems. These problems may be divided into three main categories: Hardware problems (i-iii, see below), communication problems (iv-vii), and policy problems (viii-xii). Although all problems listed below do not necessarily occur in any developing country (they do in some) it is safe to assume that in order for large-scale research support projects, such as those carried out by the World Bank, to be successful, problems of all three kinds must be dealt with effectively.

The major problems are (the order does not indicate any priority):

(i) Lack of up-to-date, functioning equipment, suited for local conditions,

(ii) inefficient and uneconomical procedures for provision of maintenance, spare parts, specialty chemicals, and other consumables,

(iii) lack of trained service technicians,

(iv) insufficient contact with other researchers and users of research in the field, both nationally and internationally,

(v) insufficient and incomplete library holdings, especially of research journals, poor dissemination habits, and lack of strong national research publications,

(vi) lack of travel funds for participation in scientific cooperation and meetings as well as lack of means for fast communication with others in the field (fax, electronic mail),

(vii) weak or lacking professional societies in individual research fields,

(viii) insufficient monitoring of research quality and efficiency, both for individual research activities and on a national level (e.g. in the form of regularly monitored output indicators) and lack of systematic research evaluations,

(ix) lack of monetary and other incentives for active research,

(x) insufficient salaries to ensure a full time effort by the researchers, and

(xi) insufficient or lacking national research policies and overdependence on outside donors.

The need for research training is overwhelming in many developing countries. There are selective and fast increasing needs for competent research trained manpower in the private sector, especially in manufacturing industries, natural product based industries, and new, fast growing environmental companies. Although the public sector in many developing countries is not expected to grow much, selected quality improvements are often badly needed. In particular, upgrading of the higher education subsector, which has the key to many new opportunities in developing countries, often makes an extensive effort in research training necessary. The teaching staff of university systems in developing countries typically has an educational background corresponding mainly to first degrees with very few Ph.D.s and Masters degrees; most universities would want to more than double the number of teachers with advanced (research) degrees. In public universities in Indonesia, for example, 6% hold Ph.D.s and 15% Masters degrees (the numbers in private universities are much lower); a doubling in the public university system alone would require 3,000 new Ph.D.s and 8,000 new Masters, at least an order of magnitude higher than the present annual production of graduates.


The core of all national efforts must be the training of young researchers and the strengthening of productivity and work habits among other researchers. In most countries this requires development of a national research training capacity in some or most fields, although regional cooperation, with a division of responsibilities for research training in different fields, often can be very useful, especially (but not exclusively) for smaller countries. Within a general framework, emphasizing human resource development, and realizing that building of research capacity takes many years, other actions and investments must be carried out in a coordinated and efficient fashion. In particular, the specific problems listed in the preceding section may be dealt with by developing country managers, in cooperation with the World Bank and donor organizations, according to the following guidelines:

(i) Lack of up-to-date, functioning equipment. Provision of equipment should be demand and not supply driven. Make equipment available only based on an evaluation of well documented applications from research teams or groups of research teams, thereby ensuring that it will be used effectively. Secure funds also for transportation and installation of equipment. Let the researchers who will be using the equipment take part in the negotiations with suppliers. Establish a central, national office for equipment procurement that can advise researchers on technical, financial, and legal matters in connection with equipment and contracts. Do not pay more than list prices for equipment unless special services are provided. Finally, make sure that funds for operation and maintenance (frequently over 10% annually of the cost of the equipment) are available for a foreseeable future. Institutional cooperation with partners in industrialized countries, that have a solid experience with the instrumentation in question, might be a very effective way of improving equipment standards (Danida, 1992).

(ii) Inefficient and uneconomical procedures for provision of maintenance, spare parts, specialty chemicals, and other consumables. Make sure that equipment is procured with a reasonable standard package of attachments and spare parts, depending on the expected use. Ensure that the equipment supplier’s warranty period does not start until the equipment is installed and is working as it should. Investigate the local availability of spare parts during the negotiations with suppliers and try to coordinate the choice of new equipment with existing equipment of a similar nature. Support the creation of user groups for specific kinds of equipment. Avoid long waiting times, customs problems, and excessive prices in connection with the purchase of specialty chemicals by centralizing purchases from research laboratories. Quantum discounts are often very high, and suppliers are more willing to stock a great variety of chemicals if a large group of users negotiate together and are able to make realistic estimates of their combined needs.

(iii) Lack of trained service technicians. Ensure that sufficient salaries or other forms of compensation are paid to qualified technicians and that they are given opportunities for upgrading of their knowledge. User groups for specific kinds of equipment may share responsibility for training of specialist technicians in their field and may, for example, ensure that no major piece of equipment is delivered without proper training in its use and maintenance for both the researchers and a (larger) group of technicians who will be responsible for technical services on similar instruments in the future. Centralize maintenance units for specialized equipment so that they can acquire the needed total expertise and facilities.

(iv) Insufficient contact with other researchers and users of research in the field, both nationally and internationally. Provide incentives for research cooperation, both within the country and internationally, e.g. by earmarking research funds for cooperative research. Provide travel grants for serious researchers who want and need to cooperate with researchers in and outside the country. Facilitate university-industry cooperation and reward researchers for producing critical reviews of their research field directed to users of research within the country. Maximize the benefits of investments in foreign training by combining it with institutional cooperation between the home institution and the foreign host. Whenever possible, support national and international institutional twinning and sandwich-type degree programs rather than one-way arrangements with foreign institutions.

(v) Insufficient and incomplete library holdings, especially of research journals, poor dissemination habits, and lack of strong national research publications. Provide individual institutions with a minimum of research journal subscriptions (without exceeding their real needs) as part of a national plan for research journal and data base subscriptions using the most appropriate information technology available. Strengthen national cooperation between research libraries and coordinate subscriptions.

Support the development of subject-specific national or regional research journals and use these both as a channel for communication with national users of research and as a stepping stone for the country’s researchers in their emerging participation in the international research communication in the field. Avoid any kind of support for non- subject specific journals or local journals without proper review procedures and nationwide circulation. When possible and needed, encourage such journals to become national quality journals through mergers with other similar journals. Reward publication in international or high quality national journals, as well as quality publications directed towards research users.

(vi) Lack of travel funds for participation in scientific cooperation and meetings as well as lack of means for fast communication with others in the field. Provide travel support for researchers attending international conferences in active roles (e.g. as speakers). If possible, attract international conferences to the country and add follow-up activities in order to benefit as much as possible from the influx of internationally known researchers. Support specialized, national workshops so that researchers in a given field may improve their knowledge of other, related activities in the country. If possible, include access to new communication opportunities such as electronic mail, fax, etc. when major research investments are made.

(vii) Weak or lacking professional societies in the respective research fields. Support the establishment and strengthening of professional societies in academic and applied fields. Such societies may function as organizers of national workshops and international meetings in their respective fields and may be responsible for the publication of national, subject-specific journals.

(viii) Insufficient monitoring of research quality and efficiency, both for individual research activities and on a national level (e.g. in the form of regularly monitored output indicators). Lack of systematic research evaluations. Monitor all supported individual research activities and evaluate the outcomes and efficiencies. Define input and -especially- output indicators both for individual projects and for larger areas of the national research. Use this information as input in national evaluations of key research fields, which may help identify both strengths and weaknesses and may help make efficient policy actions possible.

(ix) Lack of monetary and other incentives for active research. Provide incentives to active researchers by making support dependent on performance. Use monitoring and evaluation systems to provide recognition to successful researchers. Ensure that promotion schemes recognize true research quality and provide incentives for quality control and improvement, e.g. by giving higher credit for research publication in recognized international journals and for international patents than for publication in local journals and for national patents.

(x) Insufficient salaries to ensure a full time effort by the researchers. Do not provide research support to researchers who are unable to spend sufficient time on the project. If researcher salaries are so low that second jobs are necessary, add a salary component to research grants which relieves the researchers from other time-consuming jobs. Make continued salary support of the researchers dependent on their performance.

(xi) Insufficient or lacking national research policies and overdependence on outside donors. Provide assistance for establishment and strengthening of long-term national research policies, including realistic plans for research training programs. Do not accept individual programs which are inconsistent with such national policies, but make sure that they support them. Adjust the national policies regularly, based on the outcomes of continuous monitoring and evaluation activities.

Convince potential donors that the results of their efforts are only sustainable as part of a national plan and that single projects are unlikely to succeed alone, but must be carried out in coordination with other, related activities supporting the national research capacity.


Danida (1992). Evaluation report: The bilateral program for enhancement of research capacity in developing countries. Copenhagen: Denmark.

The Danish Research Academy (1993). Report on international issues in Danish Ph.D. programs. Aarhus: Denmark.

Eisemon, T.O., & Davis, C.H. (1993). Kenya: A scientific community in crisis. In J. Gaillard & V. Krishnan (Eds.) Emergence of scientific communities in the developing world. Paris: ORSTOM.

Gaillard, J. (1992). Use of publication lists to study scientific production and strategies of scientists in developing countries. Scientometrics, 23, (57).

Gaillard, J. & Thulstrup, E.W. (1993). North-South cooperation: Contributions to research capacity building. ESP Discussion Paper (forthcoming), The World Bank.

Hoffmann, R. 91993). Some reflections on science in low-income economies. PHREE Background Paper no. 71, The World Bank.

Kells, H.R. (1992). Performance indicators for higher education: A critical review with policy recommendations. PHREE Background Paper no. 56, The World Bank.

Kornhauser, A. (1992). University industry cooperation under constraints. PHREE Background Paper no. 67 , The World Bank.

Kruytbosch, C. (1993). Science and technology indicators in developed and developing countries: Availability, quality, utility. ESP Discussion Paper (forthcoming), The World Bank.

Mansfield, E. (1993). Economic returns from investments in research and training. ESP Discussion Paper, The World Bank.

Muskin, J.A. (1992). World Bank lending for science and technology. PHREE Background Paper Series no. 51R, The World Bank.

Parker, L. (1992). University industry cooperation. PHREE Background Paper Series no. 64, The World Bank.

Thulstrup, E.W. (1992). Improving the quality of university research in developing countries. PHREE Background Paper Series no. 52, The World Bank.

Thulstrup, E.W. (1992a). World Bank lending for science and technology. In Society, science, government. Warsaw: State Committee for Scientific Research.

The World Bank (1993). Scientific research and development in transition: The case of Poland. Report no. 10757-POL.

Wu, K.B. (1992). Higher education in Hong Kong. PHREE Background Paper Series no. 70, The World Bank.

Wu, K.B. (1993). Science and technology in Taiwan. ESP Discussion Paper (forthcoming), The World Bank.