Most agricultural research is conducted by the government sector, which accounted for almost three-quarters of total agricultural R&D spending and staff (Figure 5.3).
The corresponding share of the higher education and non-profit agencies were 18 and 7 percent, respectively. Private agricultural research is still quite limited in Sub-Saharan Africa. The government sector has maintained its share of total agricultural R&D spending at around three-quarters since 1971, while the university share has risen from 10 percent in 1971 to 18 percent in 2000, largely as a result of expansion in the number of universities, especially in Nigeria and Sudan (Figure 5.4).
UNESCO (1998) reports that of all regions in the world, Sub-Saharan Africa has the second lowest research and development intensity per dollar of gross domestic product (GDP) - with a gross expenditure on research and development of only 0.3 cents for every dollar of GDP (only ahead of the Arab states that spend 0.2 cents per dollar). This level is only 12 percent of that in the United States, and only 21 percent of the world average. The low level of expenditure is reflected in low output, with African scientists producing only 0.8 percent of the world's scientific publications. Africa's share of patents in the world is close to zero. But it may get even worse - UNESCO (1998) reports that Africa's share of scientific output fell by 19 percent from 1990 to 1995.
Agricultural research expenditures for a sample of 12 Sub-Saharan African countries continued to grow slowly at about 1 percent per annum during the 1990s. This was much lower than the growth observed globally (2 percent per annum) or for the developing world (3.6 percent) during the period 1991-96 (Pardey and Beintema, 2001). The growth in research staff in the 12 Sub-Saharan Africa countries during the 1990s fell to 1.7 percent per year from the much higher figures of 3.7 percent in the 1980s and 4.9 percent in the 1970s. For the past 30 years, the growth in research staff has exceeded that of expenditures, resulting in smaller expenditures per researcher. In real terms (i.e. corrected for inflation), expenditures per researcher were in 2000 only half of what they were in 1971 (Figure 5.5).
In 2000, 74 percent of the total full-time equivalent researchers in the 14-country African sample had postgraduate-level training, with close to one-quarter holding doctoral degrees. The large countries in our sample heavily influence these averages and the quality of research staff varies markedly among countries. For example, 30 percent or more of agricultural researchers in Nigeria, South Africa, Sudan, and Cote d'Ivoire held doctoral degrees while for Eritrea and Ethiopia about one-half of the researchers held only BSc degrees. A higher proportion of university staff held PhD degrees compared with staff at other agencies, a pattern that was prevalent among most of the countries in the region and which is in line with other regions such as Latin America (Beintema and Pardey, 2001). During the 1990s, the quality of research staff - measured as the share of researchers with PhD and MSc degrees - improved. In 2000, the share of researchers holding postgraduate degrees was 8 percent higher than one decade earlier, mostly a result of a relative increase in researchers trained to the MSc levels.
Many countries received considerable financial support for training their research staff, often as part of large World Bank projects or through contributions by bilateral donors or agencies such as the Rockefeller Foundation. The education level of agricultural researchers has on average improved during the past decade despite reported high turnovers of staff. Far more researchers have been trained, but never returned to the job or left the agricultural research organizations for better jobs elsewhere. Low remuneration packages are considered a major drawback to retaining and motivating highly qualified staff. When informally asked about the impact of the HIV/AIDS epidemic on staffing, it was often considered a relatively minor issue compared to losing staff for other reasons.
In 2000, 21 percent of the total FTE researchers in a 13-country African sample were female (against 17 percent in 1991), ranging from 16 percent of those holding a PhD degree to 26 percent of those holding a BSc degree as their highest level of education (Figure 5.6) (Roseboom and Beintema, 1996). The relative number of female researchers differs substantially between the various countries.
Total public spending as a percent of agricultural gross domestic product (AgGDP) is a common research investment indicator that helps place a country's agricultural R&D spending in an internationally comparable context. In 2000, the 11 countries in our sample invested US$0.76 for every US$100 of AGGDP, which is lower than the 1991 and 1981 levels (Table 5.3). These intensity ratios ranged from just 0.2 for Madagascar and Sudan to 2.6 for Kenya, 3.0 for South Africa and 4.0 for Mauritius. The latter are well above the ratio observed for most developing countries and, in the case of South Africa and Mauritius, are above the average for the developed countries as well (Figure 5.7).
The structure and sources of funding of the principal agricultural research organizations across 12 African countries is surprisingly diverse (Figure 5.8).
The average share of government funding in 2000 was about 60 percent, while donor funding represented 38 percent. The latter is much lower than the 48 percent donor share reported by Pardey and colleagues (1997)for 21 African countries (excluding South Africa) for 1991, and suggests a moderation of the high donor dependency throughout the 1990s. As a result of high donor funding, many of the main agricultural research agencies have invested significantly in their physical infrastructure, equipment, and training of staff (Figure 5.9).
Contrary to the impression the Study Panel gleaned from the consultative workshops, salaries do not seem from these data to represent an unsustainable share of the total expenditure, averaging about 42 percent. This compares favourably with the international advanced research centres of the CGIAR, with a figure of 49 percent in 2002. Hence there would seem to be an adequate availability of total operational funds from these figures, again contrary to the impressions from the consultations. However, as there are fewer international scientists in the international advanced research centres supported by their salary budget, there are far more operating dollars per scientist in the international advanced research centres than in the national agricultural research system. (Generally it appears that salaries represent a much higher share of the national research budget than the total budget. This is because donor funds are largely devoted to non-salary components. Also in recent years there has been a major reduction in salaried support staff in favour of rehiring them on casual wages, which then appear as part of operating expenses.)
In 2000, one-half of the FTE researchers in a 14-country African sample conducted crop research (Figure 5.10a). Research on livestock totaled 18 percent while forestry, fisheries, and natural resources research each accounted for 5 percent. Among the crops, fruits (12 percent) and vegetables and maize (9 percent each) were the main items followed by wheat, sugarcane and cassava (5-6 percent each). The remaining crop researchers (53 percent) focused on a wide variety of other crops (Figure 5.10b).
The 14-country sample analyses are based on Sub-Saharan countries. No similar data are available for the North-Africa region. Box 5.5 describes the North African nars based on the consultative workshop in Morocco (held jointly with AARINENA in early 2003) that provided the Study Panel with valuable information (Besri, 2004).
