Contents - Previous - Next

This is the old United Nations University website. Visit the new site at

IFPRI report


Changing food consumption patterns and price policy in West Africa

In West Africa annual per capita rice and wheat consumption rose by more than 16 kg from the early 1960s to the early 1980s, whereas millet and sorghum consumption fell by 22 kg and maize consumption increased by less than I kg over the same period. Little wheat is produced in West Africa, but rice accounted for 15% of cereals production in the early 1980s and 21% of cereals consumption. Thus there is a large and growing regional imbalance in the composition of demand and supply of cereals.

It has been tempting to conclude that consumption switches can be explained by cereal prices favourable to rice consumers. For example, the price ratio of 100% broken rice to sorghum in the world market during the first half of the 1980s was 1.4 to 1.(); in Senegal the same ratio was 1.1 to 1.0. Similar divergences between world and West African price ratios existed throughout the 1970s and continue to be observed in a number of West African countries today. Consumption switches have also occurred because rice in both West Africa and the world market has become about one-third cheaper relative to coarse grains since the first half of the 1970s.

Since 1983 IFPRI has been involved in a collaborative research network with five research institutes in West Africa in an effort to better understand the determinants and implications of major shifts in food consumption patterns away from the traditional coarse grains toward rice and wheat, which are for the most part imported. This research culminated in a policy outreach conference based on field research. It was cosponsored by IFPRI and the Institut Sénégalais de Recherches Agricoles in Dakar, 15-17 July 1987, where some 60 participants, more than half of them policy-makers or policy analysts from the region, discussed two sets of issues crucial to the substitution debate.

The first set of issues dealt with the role of price and non-price factors in supply-demand imbalances in the composition of cereals consumption. Price factors include consumer and producer prices and exchange-rate policies. Non-price factors on the demand side include urbanization, income shifts, and availability. On the supply side, they include marketing and production costs for coarse grains as compared to rice. The second set of issues dealt with the substitution implications of price, production, and trade policies in terms of comparative costs in marketing and production. It also included the critical question of equity consequences: for example, does the welfare of the urban poor suffer when rice prices rise?

Key findings of the collaborative work indicate that, at both the household and national levels, relative prices may play only a minor role in driving up West African rice consumption and that non-price factors related to such things as income, the need to eat away from home, and occupation may be more important.

Collaborative survey work reveals that urban rice consumption is especially sensitive to work patterns in the urban household. As women enter the work force, and men work away from the home, there is strong demand for staples that can be prepared quickly at low cost and that are available in roadside restaurants. Rice meets both needs. Thus, urbanization appears to dominate price factors in this context in explaining the shift to rice. In addition, available survey evidence does not support the view that only the urban wealthy are rice consumers. It may account for a significant part of the real income of the growing group of urban poor. For example, in collaborative work by the University of Ouagadougou and IFPRI, it was found that both the poorest and the richest income terciles obtained about one-third of their cereal-based calories from rice. For the poor, this accounted for one-half of their cash expenditures on cereals. In the discussion, participants constantly reiterated the importance of detailed results of this type for policy and the need for widespread replication.

It has been suggested that a possible solution to the problems associated with substitution toward rice is the creation of a regional protected zone for rice in West Africa, characterized by a high common external tariff for rice imports from outside the region. This raises a number of issues. First, would this option improve the long-run efficiency of resource use in the region? Major consuming areas for rice in West Africa tend to be close to coastal ports, whereas the major producing areas tend to be inland. Since intra-West African road transport costs for long-distance grain shipments currently average US$0.08-$0.10 per ton per kilometre, transport costs alone are often more expensive than total import costs at the point of consumption. A tariff high enough to prohibit imports would have to cover these transport costs in addition to relatively higher domestic rice production costs. Rice prices in many of the large coastal consuming areas could double over current levels that are already substantially above world prices in some cases, even allowing for overvalued exchange rates. The long-run effect of such changes on resource allocation within West Africa and the opportunity costs of reallocations is not well known. The key area of focus here should be the best way to lower unit costs of rice production.

Second, who will benefit and who will lose from rising rice prices? The urban poor are important consumers of rice, and the consumption of rice is not especially responsive to changes in its relative price. Policies that have the effect of greatly raising urban rice prices need to foresee means of decreasing the negative effect on the poor in this growing section of the West African population.

- Christopher L. Delgado and Thomas Reardon


Research perspectives

Microcomputers in food policy research

In recent years IFPRI has expanded its microeconomic research capabilities through the collection of primary data at the household level. Field studies have been undertaken in some 20 countries, generating data bases on production, input use, income, employment, consumption, and other socio-economic variables As the collection of more and detailed primary information has increased in importance in IFPRl's research, storing and computing these data have become critical. In the past, processing large volumes of the data generated by field surveys was done primarily with mainframe computers, run by highly trained personnel, located in Washington, D.C.

Recognizing the potential of microcomputer technology, IFPRI has undertaken a new approach to field data collection and processing in which microcomputers are taken to the local research institutions collaborating in the regions of study. Field data are input into microcomputers at a central location in the regions, thus improving the efficiency and speed of the initial stages of the research effort.

IFPRI research and computer staff develop the software specifically designed for the surveys and hold on-site workshops to train local users. In a number of cases requests for information on IFPRl's approach have come from other institutions in the study country. Thus, in addition to improved data validation and faster processing, use of microcomputers is contributing to enhanced local research capacities.

Currently, microcomputers are being used in Pakistan for research on food security management; in Zambia, where IFPRI is examining the effects of technological change; and in Zimbabwe on a project to assess the linkages between rural infrastructure and agricultural development. When the projects are completed, the computers will remain in the collaborating institutions.

Workshops highlight research findings

The workshop in Dakar mentioned in the Commentary was one of six meetings sponsored or cosponsored by IFPRI since May. The topics and discussions were as varied as the participants, who included policy-makers, policy analysts. and researchers from developing and developed countries. The meetings in brief:

- Trade and Macroeconomic Policies' Impact on Agriculture, Annapolis, Maryland, USA, 27-29 May. In the seven countries for which research was presented - Colombia, Peru, Argentina, Chile, Nigeria, Zaire, and the Philippines - the implicit taxation of the agricultural sector due to the overvaluation of the real exchange rate has resulted in significant production disincentives, particularly for export-crop producers. Discussions highlighted the implications for employment, income distribution, and consumption.

- Commercialization of Agriculture, in conjunction with the Kenyan National Council for Science and Technology, Nairobi, 29 June-1 July. Research presented on the effects of sugar-cane production in south Nyanza suggests that the shift from semisubsistence to commercial agriculture is associated with an increase in income and household food consumption. However, research showed little correlation between increased income and the growth and health of preschool-aged children.

- Issues in Food Security, in conjunction with the Oxford University Food Studies Group, Oxford, 7-9 July. A major point of these discussions was that if food security policy measures such as pricing and stocking are to satisfactorily deal with household food security problems, institutional arrangements that ultimately determine the microlevel consequences of these policies - marketing, input and credit supply, and employment, health, and food subsidies programmes - must also be examined.

- Rice Policy in Gambia, in conjunction with the Gambian Programme, Planning, and Monitoring Unit for the Agricultural Sector, Banjul, 21 July. Discussions of findings based on collaborative field study suggest that competition between irrigated rice and upland crop production is high because of labour bottlenecks in peak seasons. Research noted the advantages of small-scale river irrigation systems that require small investments in infrastructure and land development. Findings show the positive consumption and nutritional effects of double-crop irrigated rice.

- Trends and Prospects of Cassava in the Third World, Washington, D.C., 10-12 August. Participants indicated that there is considerable potential for increasing cassava yields and total output in the third world. They noted the potential for cassava as a livestock feed. an industrial raw material, and a generator of exchange earnings/savings. but suggested that the future scope of cassava use depends on improved post-harvest technology and product development. Cassava's importance to food security in Africa for both consumption and employment was highlighted.


New initiatives and focus at IFPRI

The appointment of three well-known development economists marks the start of major new initiatives in research and policy development at the International Food Policy Research Institute (IFPRI). Dr. Nurul Islam brings his hands-on experience in policy formation and implementation as former assistant director general of the Food and Agriculture Organization of the United Nations and deputy chairman of the Bangladesh Planning Commission. Dr. Michael Lipton, a British economist from the University of Sussex, contributes his focus on problems of equity and poverty, particularly for the most disadvantaged poor in the third world. Dr. Richard H. Sabot, of Williams College and the World Bank, a pioneer in the analysis of the effect of education on economic growth and equity, initiates a new research thrust on the relationships between educational expansion, labour allocation, rural productivity growth, and poverty alleviation.

IFPRl's growing interaction with policy-makers, at both the international and national levels, will be complemented by Islam, who has joined IFPRI as a senior research adviser to the director. Islam's distinguished career has encompassed research as well as policy formulation. His work has focused on issues of trade, food aid, economic planning, and national and international agricultural policies. Islam's experience in the policy process will add to the wealth of food policy analyses already undertaken at IFPRI.

Lipton, widely respected for his work on determinants of poverty, has established his reputation examining how power structures in developing societies have discriminated against the rural sectors, which often face government actions that reduce their share of scarce resources. In addition he has researched how credit, improved crop varieties, migration, land distribution, employment, and population structure and change interact with issues of efficiency, equity, and poverty. Lipton will direct the Food Consumption and Nutrition Policy Program.

Past IFPRI research on the linkages between agricultural development and economic growth has shown the importance of investment in physical infrastructure, particularly roads. An area of comparable importance is education, which accounts with infrastructure for the largest share of rural development expenditures by governments in the developing countries. Sabot's research has focused on the effects of educational expansion on labour productivity, economic growth, and the structure and dispersion of incomes. Sabot, whose work has also been concerned with rural-urban migration and other dimensions of the operation of labour markets, will co-coordinate research on development strategies.

"The addition of three such outstanding and highly respected scholars to our research team marks the beginning of a new direction in IFPRl's efforts," IFPRl's director, John Mellor, said. "Our work on the importance of agriculture to developing-country economies will gain in depth, and this will be particularly crucial for our continuing research on Africa, where the links among employment. income. and agriculture need to be intensively explored if spiraling poverty and malnutrition are to be reversed."

(From "IFPRI Update." Sept. 1987.)


Recent publications

Population theory Endogenous fertility

Although economic theory in many areas has become more complex, most theories on population growth and household decision-making have not progressed beyond Malthus, who contended that population would steadily increase and standards of living would decrease until parents are no longer able to feed their children. Modern theorists have incorporated Malthusian concerns about limited natural resources into growth theory, but no one has really studied how parental choices about how many children to have and how much to invest in each child's health, welfare, and education affect the well-being of present and future generations.

In Population Policy and Individual Choice: A Theoretical Investigation (IFPRI Research Report 60) Marc Nerlove, Assaf Razin, and Efraim Sadka introduce a model of family decision-making that is amazingly simple but has far-reaching implications. Because parents care about the welfare of their children, the authors contend. they will plan for their future and respond to economic constraints and opportunities in making choices affecting their children, including decisions about the number of children to have. This theory, which they call endogenous fertility, is examined to determine if it can lead to failure of the market to reach a Pareto-efficient outcome (a state of equilibrium in which no one can be made better off without making someone else worse off). They conclude that Pareto efficiency is possible with endogenous fertility because parents link their welfare to that of their children.

The market can fail to reach Pareto efficiency in several ways, however. First, parents may forget that any bequest to a child will also benefit that child's spouse and his or her parents. Second, parents' equal distribution of bequests among children may lead to an inefficient outcome because children of greater ability offer a higher rate of return on investment in human capital than can be obtained by investing in physical capital. Policies to correct these failures are formulated.

The study also indicates that the 'old-age security" theory of population growth may not necessarily be correct. This theory holds that families with no access to capital markets will try to have many children to support them in their old age. Parents may have fewer children if they can expect to obtain a larger return in terms of future consumption from capital markets than from their children. But this study points out that if parents can borrow capital to invest in more children, access to capital markets may not lead to population reductions after all.

What is the optimal size or rate of growth of population? The theory developed in this study can be applied to investigate this and many other issues related to population size and economic welfare.

Cost-effectiveness of targeted food price subsidies

For 12 months beginning in mid-1983 the Philippine government discounted the prices of rice and edible oil in villages in three areas of the country where malnutrition and poverty were severe. Other villages in the areas, acting as a control, received no subsidy. A nutrition education component was also included. This pilot food subsidy programme was carefully monitored and surveyed to determine its economic and nutritional effects, its technical and administrative feasibility, and its cost-effectiveness.

In The Pilot Food Price Subsidy Scheme in the Philippines: Its Impact on Income, Food Consumption, and Nutritional Status (IFPRI Research Report 61) Marito Garcia and Per Pinstrup-Andersen analyse the data collected from a random sample of 840 households selected from the 14 villages. They also interviewed individual members of a subsample of 140 households to determine how the additional food was distributed among household members.

The study finds that the subsidy increased both income and calorie consumption by about 9%. The calorie intake of adult males increased the most (13.4%) and that of female pre-schoolers the least (4.7%). In spite of this bias toward adults, pre-school-age children receiving the subsidy showed a marked improvement in their weight for age. Results from the nutrition education component were mixed. Nutrition education in conjunction with a subsidy seems to have had some effect, but nutrition education alone had none.

Perhaps the most interesting aspect of this report is its assessment of cost-effectiveness. The fiscal cost of transferring US$1.00 of income to the sample households through the subsidy and the cost of increasing calorie consumption by 100 calories per person per day are both calculated. Fiscal costs are estimated for five levels of possible targeting. For example, the cost of transferring $1.00 to all households in a geographically targeted area was $1.19. With the same scheme but counting only households with malnourished pre-school children, it would cost $3.61 (in 1983 prices). The cost of transferring 100 calories per person per day for a year was $3.38 under the pilot scheme, but the cost would increase to $6.83 if only households with malnourished pre-schoolers were counted.

A comparison of the Philippine pilot scheme with subsidy programmes in other countries shows that the fiscal cost of transferring income in the pilot scheme was similar to that of other schemes for which data were available, but the Philippine scheme was more cost-effective than all but one of the other programmes in delivering calories.

It was more cost-effective, first, because it was targeted geographically, thus eliminating the cost of determining which households or individuals should receive the subsidy, and, second, because its administrative costs were low. Its aim was to reach low-income households. If it had been meant only to reach malnourished children, a programme to them would have been more cost-effective.

Each household was issued a ration card for purchasing a monthly quota of 5 kg of rice and 400 g of oil per household member at a subsidized price about 30% below the market price of rice and 50% below that of cooking oil. The subsidy was distributed through neighbourhood variety (grocery) stores, and the nutrition education classes were conducted by local extension workers. The grocers could purchase the subsidized foods from either public or private wholesalers, and they were reimbursed through local banks. Because the subsidy used existing infrastructure, its administrative costs were only about 9% of the total cost. The food itself accounted for 84% and an incentive payment to retailers took 7%.

Finally, the study shows a clear relationship between low-income occupations, such as landless and tenant farm workers, and malnutrition.

Research needed to stop agricultural decline in Nepal

Nepal is among the few developing countries whose agricultural productivity (crop yield per unit of land) has declined throughout the last quarter century. In the 1970s, gross food-grain production increased by nearly 1% annually, but this was almost entirely due to an increase in cropped area rather than in productivity. Because of a 2.7% annual population increase during this period, the per capita net food-grain production dropped from 185 to 160 kg.

In Agricultural Research in Nepal: Resource Allocation, Structure, and Incentives (IFPRI Research Report 62) Ram P. Yadav traces the evolution of his country's agricultural research system and makes detailed recommendations for strengthening it. He argues that a strong research system is essential in order to effectively adopt or adapt the modern technology that Nepal must have to expand its predominantly agricultural economy. Agriculture accounts for about two-thirds of GDP and 80% of export earnings and is the main livelihood of almost 95% of the people.

Since the early 1960s, the government has allocated an increasingly larger share of the national budget to the agricultural sector, but the share of the agricultural budget designated for research has declined. Between 1970/71 and 1980/81 the research share dropped from 13.6% to 5.4% - and about two-thirds of that was actually spent on non-research production-support activities. The annual average amount spent on research from 1978/79 to 1980/81 was only 0.10% of the agricultural GDP. This compares poorly with the average 0.56% investment of 51 other developing countries in 1980. Government emphasis on nationwide food-grain production has resulted in neglect of horticulture and livestock development, which have much potential for the hill and mountain areas that constitute 78% of Nepal's land and contain nearly 60% of the population. Because of inappropriate agricultural methods, these areas have serious erosion problems that cause siltation and flooding both in the fertile plain and beyond Nepal's borders.

In interviews with 120 researchers who represent a broad range of disciplines and experience, Yadav found widespread low morale and inefficiency. Complaints included unclear research policies and programmes, poor guidance and supervision, inadequate training and facilities, little opportunity for publication, and frequent transfers. Frequent transfers of personnel are also a serious problem at high levels of the government. In the 16-year period from 1967 to 1983, there were 16 ministers of agriculture; a one- or two-year term is typical for most senior positions where agricultural policy and priorities are established.

The study finds that the foremost requirements for reform of the present agricultural research system are establishment of a high priority and an adequate budget for research activities, unification of all these activities under one autonomous organization, a change in the focus of foreign aid from short-term production programmes to long-term research and development, and closer collaboration with other national and international research systems.

Panterritorial pricing policies

In July 1974 the government of Tanzania announced a panterritorial pricing policy placing a uniform producer price on certain crops, regardless of where in the country a crop was grown or how high the cost to transport it to market. This decision was part of an effort to equalize development among the country's regions. But severe climatic conditions, high oil prices, and a disastrous maize harvest forced the government to import large quantities of food at the same time that it had to absorb the transport costs from the new policy, leading to severe food shortages and balance-of-payments problems.

In Effects of Panterritorial Pricing Policy for Maize in Tanzania, Yuriko Suzuki and Andrew Bernard examine the effects of the policy on production and marketing and alternative means of arresting its increasing financial burden. The study identifies surplus and deficit areas, taking into account the existence of parallel markets.

Although many African countries have adopted panterritorial pricing policies, the results were more severe in Tanzania because the most suitable regions for maize production are far from population centres. Suzuki and Bernard suggest that regions near the southern border could continue to grow maize and export the surplus to neighbouring countries, thus saving on transport. The foreign exchange earned could then be used to purchase imports for the capital area. This, however, could cause donor countries to curtail aid on the assumption that a country that is exporting staple foods must have a surplus. This study concludes that donor countries need to revise their thinking so that deficit countries such as Tanzania can export grain from far-flung growing regions when that is more cost-effective.

Dairy development in India

Cooperatives and the Commercialization of Milk Production in India: A Literature Review, an IFPRI working paper by Harold Alderman, George Mergos, and Roger Slade, reviews the history of the Indian dairy co-operative movement. The authors concentrate on Operation Flood, an ambitious government programme that has created thousands of village milk producers' co-operatives grouped into unions that build, own, and operate processing, collection, and distribution networks throughout India. The effectiveness of Operation Flood is debated among scholars, some of whom see social ills or inappropriate priorities such as a shift of milk consumption from rural to urban consumers and diversion of food grains to animal feed, while others claim positive results such as increased employment for rural households and better nutrition for both rural and urban low-income groups.

The evidence suggests that the primary determinant of increased milk production is an increased percentage of buffalo in the national herd. Availability of feed appears to be a major constraint to dairy development, but cultivation of fodder crops is not popular among farmers.

Although empirical data are limited, comparative studies indicate improved nutrition linked to increased income from dairy development. The authors recommend studies of relative and absolute income changes rather than measurement of milk consumption to determine welfare impacts.

Contents - Previous - Next