2.3 The developing countries: magnitude of the food problem and historical developments

Contents - Previous - Next

The overall picture

How big is the food availability problem in the developing countries? How has it developed over time? How many people suffer from undernutrition? These are some of the questions addressed in what follows. First, the magnitude of the problem and its evolution over time. Per caput food supplies in the developing countries as a whole have been increasing, from 1950 calories in the early 1960s to 2470 calories in 1988/90 (revised and more recent data to 1992 are given in the Appendix 3). This happened while their population grew from 2.1 billion to nearly 4.0 billion. Therefore, significant progress has been made. This progress, or lack of it in some cases, can be seen visually in Figure 2.4 where the evolution is shown for the individual regions.

The result of these developments has been that in 1988/90 only some 330 million people, or some 8.5 percent of the developing countries' population, lived in countries where per caput food supplies are extremely low-under 2100 calories (Table 2.2). Thirty years ago these numbers were 1.7 billion or 80 percent of the total. The progress achieved can also be seen by looking at the picture from the other end. Some 650 million people, or 17 percent of the total population of the developing countries, lived in 1988/90 in countries with per caput food supplies over 2700 calories. Again, 30 years ago these numbers were only 35 million or under 2 percent of the total.

It is obvious, however, that, impressive as this progress has been, it has not been fast enough nor of a pattern that would have raised per caput food supplies in all countries to levels usually associated with significant reductions in the population suffering from serious problems of food insecurity and undernutrition. What is the size of the population in this category depends, of course, not only on national average per caput food supplies but also on how equally such supplies are distributed within each country. The empirical evidence here is scant. However, from the few countries with "good" quality distributional data it is found that on average (i.e. taking the simple average of high-inequality and low-inequality countries) the proportion of the population undernourished is around 10 percent when per caput food supplies are around 2700 calories. It is typically in the range 15-35 percent when national average calories are in the range 2200-2500. The data in Table 2.2 indicate that some 3.3 billion people live in countries with under 2700 calories and some 2 billion in countries with under 2500 calories.

The above numbers give an idea of the magnitude of the problem, if progress were to be measured in terms of national average per caput food supplies. However, not all people in countries with low national averages, even very low

Table 2.2 Population living in developing countries* with given per caput food supplies, 1961/63 to 1988/90

Per caput food supplies (car/day)
  Three-year averages Population (million)
Cal/day 1961/63 1969/71 1979/81 1988/90 1962 1970 1980 1989
Developing countries
Under 2000 1810 1960 1900 1784 1581 1046 197 123
2000-2100 2045 2030 2100 2040 114 746 771 211
2100 2300 2210 2200 2180 2225 176 338 483 1425**
2300 2500 2400 2395 2345 2405 196 230 1234 212
2500 2700 2655 2560 2655 2635 38 176 103 1327
2700 3000 2785 2870 2780 2795 14 55 293 286
Over 3000 3075 3265 3070 3120 21 24 190 365
Total 1950 2120 2330 2470 2141 2616 3271 3950
countries 3030 3200 3290 3400 989 1074 1168 1242
World 2290 2430 2580 2700 3130 3690 4439 5192

*All countries with food balance sheet (FBS) data. The data in this table and Table 2 3 are before the most recent (1994) revision of the FAO FBS data and the 1992 revision of the population data. lncludes China (pop. 663 m.) and India (pop. 462m.).

China (pop. 816m.).
India (pop. 555 m.).
India (pop. 689m.). China (pop. 978m.).
**lndia (pop. 836m.). China (pop. 1102m.).

Ones, are subject to undernutrition. And there are undernourished people even in countries with relatively high national averages. Therefore, a more appropriate estimate of the incidence of undernutrition can be obtained from a combination of the national average food supplies with a distributional parameter and some notion of a nutritional threshold level, i.e. a level of food intake below which a person chronically subjected to it can be classified as undernourished. This method and data (supplemented in many cases by educated guesses for the value of the distributional parameter and the shape of the statistical distribution curve) have been used by FAO to derive rough estimates of the numbers of persons in the developing countries which can be classified as undernourished (FAO, 1992a).

The estimates of the incidence of chronic undernutrition thus obtained were most recently published on the occasion of the International Conference on Nutrition (December 1992). They are shown in Table 2.3. Progress has been made, but the magnitude of the problem remains significant. The largest, though declining, numbers are to be found in Asia but those in sub-Saharan

Table 2.3 Estimates of chronic undernutrition in the 93 developing countries of the study

  Year Per caput food supplies (car/day) Total population
% of total population Million
Africa (sub-Sahara) 1969/71 2140 268 35 94
1979/81 2120 358 36 129
1988/90 2100 473 37 175
Near East/
North Africa
1969/71 2380 178 24 42
1979/81 2830 233 10 23
1988/90 3010 297 8 24
East Asia 1969/71 2020 1147 44 506
1979/81 2340 1392 26 366
1988/90 2600 1598 16 258
South Asia 1969/71 2040 711 34 245
1979/81 2090 892 31 278
1988/90 2220 1103 24 265
Latin America and Caribbean 1969/71 2500 281 19 54
1979/81 2690 357 13 47
1988/90 2690 433 13 59
Total 1969/71 2120 2585 36 941
1979/81 2320 3232 26 843
1988/90 2470 3905 20 781

Persons who, on average during the course of a year, are estimated to have food consumption levels below those required to maintain body weight and support light activity. This threshold level (ranging from an average of 1760 car/person/day for Asia to 1985 for Latin America) is set equal to 154 times the basal metabolic rate. For more explanations see FAO (1992a)

Africa have been increasing rapidly, both in absolute terms and as a proportion of the region's total population.

The above two alternative approaches to gauging the magnitude of the food problem in the developing countries (population living in countries with given per caput food supplies, Table 2.2, or numbers of undernourished, Table 2.3) provide useful starting points for looking into prospects for the future which are addressed in Chapter 3. That is, one would need to speculate about how average per caput food supplies may develop in the different countries and what may happen to within-country distributions. Unfortunately, the historical data and ability to speculate on prospective changes in socioeconomic and political structures having a bearing on within-country inequalities allow little scope for saying much on the distributional aspect. It is a somewhat less arduous task to address the question how the per caput food availabilities may develop in the future in the different countries. For this purpose, the empirical evidence is analysed in order to, as far as possible, understand what have been the factors responsible for the fact that some countries have made progress in raising per caput food supplies and others have not, or have experienced outright deterioration. One way of addressing this issue is to examine the historical developments in those countries which performed well in the past; and to do the same for those countries which failed to make any significant gains or experienced outright declines (the following discussion draws on Alexandratos, 1992).

The correlates of success

For the purposes of this discussion, the group of countries which performed well in the historical period includes those which, starting from low or very low levels of per caput food supplies 30 years ago (between 1650 and 2300 calories in 1961/63) had reached by the late 1980s high or medium-high levels (between 2600 and 3300 calories in 1988/90).7 The main characteristics of their historical evolution which probably explain much of their progress in raising per caput food supplies may be summarized as follows:

1. All of them had above average economic growth rates, as evidenced by the growth rates in their per caput incomes. This seems to be the most prevalent common characteristic of these countries.

2. In most countries, there was a spurt in the growth of food imports, particularly in the period of rapid gains in per caput food supplies, as evidenced by the increases in per caput net imports of cereals. This meant rapid declines in their cereals self-sufficiency. But there were exceptions. In particular, China and Indonesia did not follow this pattern as their own agricultures grew to provide the additional food supplies and, most probably, was a key factor in raising per caput incomes.

3. A contributing factor to the nutritional improvement in this group of countries has been the fact that global agriculture provided readily and without much strain the food imports that underpinned the growth of their consumption, mostly in the 1970s. It is noted, however, that the historical experience of responsiveness of world agricultural production to increases in demand must be interpreted with care, in particular for drawing conclusions about the potential of world agriculture to respond to spurts in demand in the future. This is because the data showing that increasing quantities were forthcoming in the world markets at the same time as prices declined are vitiated by the agricultural support and protection policies of major countries (e.g. USA, EC, Japan). In practice, part of the costs of delivering the increased output were covered by the heavy subsidies granted to agriculture in these countries (see Note 4). It is not known how world agricultural production would have reacted to the increasing demand in the absence of such distortions. Some indications can be obtained from the trade liberalization studies which simulate world food markets with removal of such distortions. They generally indicate that world food prices would have been somewhat, but not much, higher (Goldin and Knudsen, 1990).

4. There emerges a mixed picture concerning the role of domestic agricultural growth in the process of increasing per caput food supplies. All sorts of situations are encountered, with per caput production declining in some countries while increasing in others at moderate or very high growth rates. Drawing the conclusion that agricultural growth does not matter would not be warranted, however. More likely, agricultural growth plays a largely subsidiary role in countries where agriculture is a small sector in the economy with a relatively small share of the population depending on it for a living (see Section 2.2), and much of the gains in economic growth and import capacity derive from the nonagricultural sector, particularly from non-agricultural commodity sectors. This seems to have been the case of many of the oil-exporting countries. But in countries where such conditions are not prevalent, agricultural growth seems to be an essential ingredient in the process of increasing per caput food supplies, through its role in the provision of supplies, income and employment, and in support of economic growth and the balance of payments. China's experience in the post-reform period after 1978 seems to conform to this pattern.

5. In all countries, much of the quantum improvement in per caput food supplies was achieved in a relatively short period of time, in most cases around 10 years. Judging the durability of such gains is more difficult. There are examples of countries where improvement and retrogression of per caput food supplies follow the commodity boom and bust cycles. It is, therefore, possible that the food and nutrition gains will tend to prove more durable in countries in which the circumstances that brought them about are part and parcel of wider economic and social transformations, e.g. China and Korea (Republic). The same probably holds for countries in which the windfalls from commodity booms are put to good use to bring about such transformations, as indeed happened in some of the countries in the group analysed here.

6. Finally, the relationship between the growth rate of population and that of per caput food supplies does not produce any consistent pattern. Fast progress in the latter variable occurred in countries with very high rates of population growth, e.g. Libya and Saudi Arabia. Again, no hasty conclusions may be drawn. In particular the evidence from these countries may not be taken to mean that fast demographic growth is not an obstacle to improving welfare. All countries in this group with high population growth rates experienced special circumstances (the windfall income gains from the oil sector) and indeed in some cases the high demographic growth was partly the result of such special circumstances, e.g. due to labour immigration. The proposition that poor countries with high population growth face a more arduous task in improving welfare than those with lower population growth is in no way falsified by the experience of the above countries.

The correlates of failure and retrogression

At the other extreme, the study of the experiences of the many countries which, starting from low initial conditions 30 years ago, failed to make progress or suffered outright declines, should provide some insights as to the reasons for failure. The study of the relevant data for a sample of these countries leads to the following conclusions.

For the great majority of these countries one could have predicted that the food situation would be really bad even before looking at the data. Many of them are in sub-Saharan Africa, a fact which by itself tells a lot, given the overall economic and agricultural stagnation that has been plaguing the region for some time now. Add to this the fact that many of these countries, both in Africa and elsewhere, have suffered or are still going through severe disruptions caused by war and political disturbances and one has in a nutshell the explanation for failure and retrogression on the food and nutrition front.

The data do no more than confirm this impressionistic prediction. Indeed, the most common characteristics of these countries are declines in per caput incomes and per caput agricultural production. The two are not, of course, independent of each other. Their per caput food imports did increase, often by means of food aid. However, in contrast to the experiences of the countries in the preceding category, their per caput imports of cereals remained at generally modest levels, while the declines in cereals self-sufficiency were accordingly contained, at the cost, of course, of stagnant or declining per caput food supplies.

Some generalizations and how they can aid the assessment of the future

The preceding discussion, based as it is on an impressionistic examination of the data for a restricted number of countries, is meant to provide some clues as to what have been the main correlates of success or failure in raising per caput food supplies in the different countries. However, it is far from a complete analysis. At best it indicates that success is commonly associated with sustained growth in per caput incomes and with varying combinations of growth in domestic agricultural production and import capacity.

A more complete and formal analysis of the data for all developing countries (similar to that reported in Alexandratos, 1992) confirms that these findings hold in a general way, meaning that these three variables alone (per caput incomes, growth of agriculture, food imports) explain only partly the differentials among countries of per caput food supplies. For example, for the 65 net cereal-importing developing countries thus analysed only 25 had per caput food supplies within a range + 5 percent of those justified (or predicted) by their levels of the above three variables. Another 24 countries were within a range of between +5 and 10 percent. The deviations for the remaining 16 countries were outside the + 10 percent range. These findings suggest that some countries achieve relatively "satisfactory" levels of per caput food supplies at comparatively low levels of per caput income. And others have per caput food supplies lower than their per caput incomes would lead one to expect. Understanding (or hypothesizing about) the reasons for these discrepancies is probably the most important insight for guiding the quest for policy responses to the problems of food and undernutrition. That is, if some countries have managed to make progress on the food front while remaining essentially poor on the overall income side, would not other countries with a food problem learn something from their experience?

Obviously, other factors beside per caput incomes, growth of production and food imports, such as food prices and the distribution of income or the incidence of poverty, are important determinants of the inter-country differences in per caput food availabilities. Moreover, some of the differences are explained by the fact that the conventional income statistics, expressed in a common currency, usually the dollar, can distort the relative positions of the different countries on the income scale because the internal purchasing power of this somewhat fictitious dollar can vary widely among countries. For example, Egypt, Honduras, Bolivia and Zimbabwe with reported per caput incomes in the range of $580-650 have per caput food supplies of calories 3310, 2210, 2010 and 2260, respectively. These differences are partly explained by the wide disparities in the purchasing power of the dollar in the different countries. When incomes are corrected for this factor and converted to dollars reflecting purchasing power parities, they become $3600 (Egypt), $1820 (Honduras), $2170 (Bolivia) and $2160 (Zimbabwe) (income data from World Bank, 1993a, Tables 1, 30). In addition, public policy which influences access to food, directly (e.g. through public food distribution schemes) or indirectly (e.g. through policies to alleviate poverty), can be an important factor in explaining the observed differences in per caput food supplies among countries.

Important as these other factors are, they are difficult to account for systematically in the assessment of the possible evolution of per caput food supplies in the future. The reason is that not enough is known, nor can it be deduced in any valid and systematic way, about how such things as the income distribution, the incidence of poverty and public policies affecting access to food may evolve in the individual countries, let alone the prospects that peaceful conditions will prevail and when in the countries plagued by war and other disturbances. This being so, the main guide for the projections is the effects of growth in per caput incomes (mostly obtained from other organizations), the evaluation of the agricultural production prospects undertaken for this study and a notion of possible levels of food imports. Yet, the preceding discussion indicated that these three variables alone are not very good predictors of inter-country differences in per caput food supplies and that other country-specific factors must be taken into account. This apparent contradiction is bypassed by starting each country's projections of per caput food supplies from its own base year figure. The latter already embodies the effects on per caput food supplies of these "other" country-specific factors as well as the effects of per caput incomes, the growth rate of agriculture and the level of net food imports. Notionally, ignoring these "other" factors means that they are assumed to continue to play in the future the same role as they played in determining the base year food supply levels. And, naturally, some of these factors, in particular the prevalence, or otherwise, of peace, are supposedly already incorporated in the income growth projections of other organizations which, as noted, are taken as exogenously given in the projections of per caput food supplies, as well as for providing the general overall economic background within which the production and food import prospects are evaluated (see World Bank, 1994a, p. 24). For example, it is difficult to assume that countries projected to be on a low overall economic growth trajectory can mobilize the significant amounts of resources for investment in agriculture, rural infrastructure and human capital needed to underpin significant accelerations in agricultural production.

Contents - Previous - Next