Journal Issue

Population: exploring the food-fertility link: How can more and better food affect the rate of population growth?

International Monetary Fund. External Relations Dept.
Published Date:
December 1975
  • ShareShare
Show Summary Details

Shahid Javed Burki and Shahid Yusuf

The food crises which swept through Africa and Asia in 1972–73 caught the world community unprepared; massive food shortfalls, accompanied by burgeoning population—especially in the developing countries—have become the subject of countless debates and discussions in international forums. This situation has posed a new challenge for planners: how to achieve a balance between food production and the number of people inhabiting the earth.

At the present annual growth rate of about 2 per cent, world population will rise to about 5.3 billion by the year 1990. Meanwhile, without a massive investment in the agricultural sector, world foodgrain production will be about 1.6 billion tons in 1990 while expected demand at that time will be about 1.9 billion tons. The world, and the developing countries particularly, could be said to be facing the Malthusian specter once again. The basic logic in neo-Malthusianism is hard to dispute; there is no question that there is a limit to the earth’s capacity to produce “conventional food” and that the time it will take to reach this limit is dependent on the rate of population growth. (See Tables 1 and 2.)

Table 1.World population, 1930–2000

(In millions)

PopulationYears required to double population at same average annual growth rate as in 1960-70
YearWorldDeveloped countriesDeveloping countriesWorldDeveloped countriesDeveloping countries
Sources: United Nations Demographic Trends In the World, 1950-1970. E/Conl. 60/BP/1 (May 3, 1973), Table 1, p. 4; and unpublished population projections of the United Nations (New York, 1974).
Sources: United Nations Demographic Trends In the World, 1950-1970. E/Conl. 60/BP/1 (May 3, 1973), Table 1, p. 4; and unpublished population projections of the United Nations (New York, 1974).
Table 2.Cereals: Actual consumption and projected trend demand by main types of utilization, 1970–90
Actual consumptionProjected demandPercentage IncreaseRates of growth
(Million metric tons)(Percent)(Per cent annually, compound)
Developed countries
Other uses84.9100.6109.5116.418.529.
Per capita (kg.)576.0623.0649.0663.06.212.715.
Developing market economies
Other uses46.464.175.488.538.162.590.
Per capita (kg.)220.0233.0240.0246.
Asian centrally planned economies
Other uses24.632.636.039.132.546.358.
Per capita (kg.)257.0290.0298.0304.012.816.
Other uses155.9197.3220.9244.026.641.756.
Per capita (kg.)333.0349.0355.0357.
Source: Estimates by the Food and Agriculture Organization.
Source: Estimates by the Food and Agriculture Organization.

Current economic models for analyzing demographic phenomena have long lost their vitality and their capacity to provoke fresh insights. But the problems facing the majority of the world’s populations need immediate analysis, so that policymakers can begin the task of engineering solutions for the food-and-population problem.

One way of attempting to solve this problem is to establish a clear relationship between food and family size. However, defining an unambiguous relationship between these variables is possible only when we have perfect knowledge about all the intervening variables. This may not be possible, at least for the present, for two reasons. First, in order to establish a link between food and fertility we have to explore the nature of a number of other subsidiary relationships. Very little is known about them; psychologists, sociologists, anthropologists, and followers of other disciplines have only recently begun to interest themselves in these relationships. Second, it is difficult to be absolutely certain about the nature of human response. Even when all the elements in human environment have been specified with a great deal of accuracy, actual human reaction may still deviate considerably from the anticipated response.

However, sufficient information has been gathered about the relationship between various social, economic, sociological, and physiological factors to suggest a way in which food can be seen to influence fertility. We can therefore spell out the various elements in this equation, which will help us to draw some policy conclusions.

The food-fertility equation

In the family-forming behavior of rational economic man, fertility appears as a variable dependent on a number of factors that shape his environment. Two elements in the food-fertility equation deserve special mention: income or “child services” required by the family at a particular period in its life cycle, and the stock or number of children needed to produce them. The first element in this equation is provided by an extension of the economic theory of fertility according to which parents desire children for the benefits they generate. Implicit in this “child services” approach is the assumption that the desired amount of services can be provided only by a given number of children or, alternatively, by children of a given quality. Therefore, not only is there an optimum family size determined by family circumstances, but there is also substitutability between the number of children and their quality. A family would be willing to trade off a larger number of children of poor quality against a few of higher quality.

Therefore, each individual is considered to have defined an optimal sized family unit and this definition requires that he achieve the target size within a given segment of the family life cycle. We can look upon the process through which the family unit moves toward this optimum size as a form of lagged adjustment. The length of the lag is determined principally by fertility and child mortality. The individual decisionmaker adjusts the rate of accumulation of offspring within the family in accordance with a time schedule. If the children are to serve as economic assets, they must have reached a certain age by the time the head of the family is ready to relinquish his economic burdens; hence not only is it imperative that the desired stock of children with the appropriate ratio of males and females be attained within the allotted time, but it is also important that adequate precautions be taken to prevent their numbers from dropping to a suboptimal level at a later stage in the family life cycle. At that point the age of the mother may greatly reduce the chances of having more children, and, even if they are conceived, their value as economic assets would be marginal; more likely they would be a drain on the resources of the family unit at a critical time in its life cycle.

Numbers and quality

Thus the head of the family must decide not only the speed at which the optimal number is approached but also the extent of overshooting so as to have a reserve which would compensate for wastage through mortality. The latter calls for some rough calculations and in fact is enshrined in societal traditions. It is these traditions which indicate to the decisionmaker what the safe degree of overshooting happens to be, given past experience with “age specific mortality” (the frequency of death at a given age). According to this argument, families deliberately overshoot, producing a stock of children greater than this requirement. A balance between the number of children and the services provided by them would only be reached if the age specific mortality rate built into the family’s fertility equation actually materializes. This may not happen; in situations of declining mortality rate, most families are likely to be overstocked with children. Overshooting of family size targets therefore is at the bottom of the population burden borne by developing societies.

Thus by reducing the rate of infant and child mortality and by informing couples of the extent of change that has taken place we should expect a decline in birth rate that would be proportional to the decline in mortality.

The quality of the stock of children also affects family size. The quality-quantity trade-off among children has been used to explain the decline in family size that has taken place in Western countries. According to this hypothesis, as incomes grow, elasticity tends to favor quality over quantity. This relationship provides us with one other policy instrument for affecting family size. By bringing about an improvement in the quality of children and by informing the parents of the extent of this improvement, we should, once again, expect a decline in the birth rate. Therefore, decline in mortality and investment in human capital become critical factors in reducing fertility.

“Recent mortality reduction in the developing countries has been achieved principally through low-cost, efficient public health, and disease control measures, which have been largely independent of economic development,” says a recent staff study of the World Bank (Population Policies and Economic Development, Johns Hopkins University Press, 1974). “For example, significant mortality declines have occurred both in Mexico, which has experienced rapid economic growth, and in Sri Lanka, which is a country with relatively slow economic growth.”

The study goes on to say that “although disease control will undoubtedly be still further expanded, much of the future progress in reducing mortality will depend on improving socio-economic conditions which govern nutritional standards, environmental sanitation, and medical care systems.”

Nutrition and family size

Improvements in nutrition affect the process of family formation in two ways: they alter the size of the family, as well as the quality of the stock of children. Improved nutrition reduces the rate of infant and child mortality; with a lower risk of child loss, the families would be inclined to reduce the number of births. A higher rate of survival also affects the spacing of children. The uncertainty induced by high child mortality rates results in women being forced to bear children through most of their fertile period. But when mortality declines, family formation can be concentrated in an earlier stage of family life. However, a quick building of the necessary stock of children is possible only when mothers receive an adequate amount of nutrition. Reducing the space between births without increasing nutrition for the mother is likely to increase infant mortality.

In formal terms, better nutrition can be viewed as an outward shift of the family production function, leading to more surviving children. Once the family becomes aware of this shift, there is considerable probability of a fall in the number of children desired by the family. Ultimately, this calls for a revision in the implicit cost-benefit calculation by which the family arrives at the figure for the optimal number of children. But before this adjustment can take place, not only must the decisionmakers (both parents) have full access to the information which shows the link between nutrition and human productivity, but also this information must be validated either at the personal level or through firmly incorporated social traditions.

Evidence of food-fertility link

Recent research on fertility behavior in developed and developing countries, among nomads and in sedentary peasant societies, among urban and rural people, and in different economic and social classes, provides us with useful evidence to suggest that the various interrelationships in the food-fertility link may in fact be valid.

Of the two critical elements identified in the fertility equation, the first—use of child services—is difficult to quantify. Econometric validation of this hypothesis does not yield clear results for the simple reason that the economic theory of fertility cannot be applied to the developing countries without significant modification. It is possible that in a developing country parents’ demand for children might outstrip their ability to produce them. When a family has so many children that the net economic and psychic cost of an additional child exceeds the economic and psychic cost of fertility control, the economic theory of fertility once again becomes operative. Below this threshold we do not find clear evidence to show that the theory holds in the case of developing countries.

However, there is a positive impact of infant and child mortality on fertility. Evidence from a number of countries indicates that parents correct for actual as well as anticipated wastage. In a study based on a fertility survey in Cairo (see Abdel R. Omran, Egypt: Population Problems and Prospects, University of North Carolina Press, 1973), it was found that “those with no experience of child death reproduced to a number that just exceeded their expectations—perhaps to safeguard against expected loss. Women with experience of child death continued to reproduce until they achieved a comparable margin, but at a higher level.”

Demographers have also made the not very surprising discovery that female literacy is inversely correlated with birth rate. The more obvious explanation uses the opportunity cost of time as the determining factor: it is costly for literate women with higher permanent incomes, who could otherwise be taking part in productive activity, to bear and raise children. These women also desire a higher standard of education for their offspring. The cost of children is therefore higher in more literate families and may increase exponentially with the number of children.

This explanation, while valid, does not fully take into account the role played by education. Literacy is critical, not only because it raises the opportunity cost of female time, but also because it helps the family to comprehend more fully the changes that have occurred or are occurring in its environment.

“the South Asian subcontinent may have to accommodate 1.45 billion people by the year 2000”

Transmitting information to families

In the case of the perceptual lag which slows down the individual’s response to a decline in child mortality, the information is available but it has not been translated into what we may call effective information units (e.i.u.’s) to affect decisionmaking. The speed with which information (in this case about decline in the rate of fertility) can be transmitted depends upon the ease with which accuracy of the information can be verified by the family decisionmakers.

Therefore, the productivity and the utility of a unit of information is related to a varying extent with its verifiability. An individual’s ability to verify information, and hence improve his decisionmaking efficiency, depends on his accumulated knowledge and the institutionalized procedures which have been evolved to communicate information and ensure its reliability. The accretion of knowledge is related to the level of education, while the nature of institutions is related to the strength of contractual relations within the society and the complexity of organizational forms.

It is important to realize that without some organization and institutional framework, there cannot be an easy flow of e.i.u.’s. Unless the organization can be created to knit together the disparate social units, the system will not be able to retain new information and hence will not be able to rise to a higher level of technological sophistication (exemplified by birth control) and economic modernity (exemplified by smaller families). Thus the lack of organization not only reduces the search for additional information but also affects the usage of existing information. In particular, this is due to the distortions which creep in when measures taken to retain knowledge over time lead to a partial loss of message content.

The difficulties facing the developing countries are obvious. In order to expedite the modernization of a society, it is necessary for the governing and planning body to funnel a great deal of information to the target people, who frequently have no means of decoding and verifying these messages. Even more frustrating is the lack of organizations to serve as intermediaries in the process of information transfer. Under the circumstances, it is not adequate simply to educate the masses, although this is a first step in breaking the communication deadlock. What is much more essential is the creation of a reliable and dynamic organizational framework.

Though nutrition occupies a central place in the fertility equation, the full impact of better nutrition on family forming behavior can only be felt when its influence on mortality, child and mother’s health, and child spacing can be perceived. The concept of effective information units, which are dependent on literacy as well as institutions of modernization may help us explore the full meaning of the role of perception. In sum, while food is a strong determinant of fertility, to produce the necessary impact it has to operate by way of education as well as institutionalized behavior patterns.

Policy implications of the equation

The equation we have spelled out leads us to some important development policy issues. To focus on them, we can quote the case of South Asia, an area where the food-fertility equation has been under serious examination.

The South Asian demographic picture is by now a familiar one. The combined population of the subcontinent (India, Bangladesh, Pakistan, Sri Lanka, and Nepal) for mid-1975 is 783 million. Growing at a rate of 2.5 per cent a year, it could cross the 1 billion mark in mid-1985. If there is no decrease in the region’s population growth in the next quarter century, the South Asian subcontinent may have to accommodate 1.45 billion people by the year 2000.

South Asian foodgrain production in 1975 is expected to be of the order of 150 million tons, or 192 kg. per person. This is nearly 45 per cent of the per capita global average of 347 kg. Grain consumption is 202 kg. per capita, or about 5 per cent more than production. This leaves a deficit of about 8 million tons that has to be met from imports, which at current prices cost the region some $1.5 billion, or $2 per capita. This is a fairly significant drain of resources when viewed against per capita incomes that average around $102 in 1975.

South Asia therefore has a population and a food problem. With production almost stagnant, the region faced famine conditions in 1974. Unless the two problems are tackled effectively, the region will remain susceptible to recurrent famines.

There is one element of development strategy in our fertility-food equation that may be the most appropriate for South Asia—the productivity of small farmers.

Increasing productivity

The Green Revolution in South Asia, ushered in with such fanfare during the mid-1960s, now seems to be at a standstill. It succeeded in increasing the food-grain output of South Asia from 105 million tons in 1966 to 148 million tons in 1970. While bad weather was partly to be blamed, there is now considerable evidence to show that the rate of adoption of new technology has slowed down in recent years. It has slowed down because the middle and large farmers, who were the first to be affected by the Green Revolution, have now fully switched to the new varieties. In doing so, they have created a fairly wide productivity margin between themselves and the small farmers. The land productivity of large farms is about 40 per cent more than that of small holdings (Table 3). If this productivity differential is closed, significant additional output of foodgrains could be obtained from South Asia. Using data from India, it can be estimated that small farmers could produce an additional 20 million tons in the ten-year period between 1975-1985 at a cost of $8 billion (in constant 1974 prices). This implies a very favorable incremental capital output ratio (ICOR) of 2.2. The same evidence suggests a higher ICOR for middle and large farms of the order of 4.0. It is therefore economically efficient to look to the small farmers for increasing the food-grain output of South Asia.

Table 3.Productivity differential according to farm size in South Asia


Farm sizeNumber of farmsFoodgrain areaYieldOutput
(Hectares)(Million)(Million hectares)(Kg/ha)(Million)
Less than 126.68.68137.0
More than 102.635.686230.7
Note: Production data include milled rice.

Farmers with leas then 5 hectares are “small farmers”

Middle farmers.

Note: Production data include milled rice.

Farmers with leas then 5 hectares are “small farmers”

Middle farmers.

By breaking down the demographic data for all South Asia by regions, we begin to notice some very interesting divergences. Interregional comparisons between crude birth and death rates suggest some correspondence between agricultural output and these variables. In regions where agricultural advance has been rapid (for instance, the two Punjabs) the rate of fertility, as well as of infant and child mortality are considerably below the areas that have not experienced high output gains (for instance the States of Bihar and Assam in India, and the Province of Baluchistan in Pakistan). Within regions, some recent studies have suggested significant fertility differentials between economic classes (high-, middle-, and low-income groups), occupational categories (urban professionals, farmers, tenants, and landless workers) and castes (Brahmins, Untouchables, and so on). These studies seem to confirm what has now become conventional wisdom about changes in fertility—that declines in fertility rates take place with increases in income and improvements in social status.

This takes us back to the food-fertility equation: by increasing the food output and the food consumption of higher fertility groups, we should be able to produce a decline in their birth rate. This appears to be the most appropriate way’ of solving the two problems of food and fertility. It is up to the policymakers to marshal the organizational resources at their disposal for the provision of information, the provision of credit to procure inputs, and the provision of some degree of economic security so that the small farmers can make the transition to a new set of norms for family size after having shifted irreversibly to a higher production level. The main problem lies in organizational inadequacy; it remains the biggest impediment to the flow of information. But the most serious shortcoming by far is that the organizations that exist are not attuned to cultural differences and lack a coherent strategy for institutional modification to maintain cultural continuity.

Other Resources Citing This Publication