Contents - Previous - Next
This is the old United Nations University website. Visit the new site at http://unu.edu
International Food Policy Research Institute, Washington, D.C., USA
During the past 10 to 15 years, many developing countries have experienced severe economic crises due to rapid changes in oil prices, falling and unstable prices of many of their important export commodities, rapidly increasing rates of interest, increasing dependence on foreign borrowing, and a series of adverse domestic developments such as rapidly increasing government deficits, certain price distortions, and low returns on investments due to structural problems in some private sectors of production and inefficiencies in the public sector. The crises are reflected in reduced foreign exchange reserves, increasing needs for foreign borrowing, and inability to meet foreign-exchange obligations such as debt servicing.
In attempts to deal with the crises, countries have implemented a variety of policies, here referred to as macroeconomic adjustment policies. These policies include devaluation of domestic currencies, reductions in price distortions and government spending, and many other measures as discussed below. While they are aimed at long-term solutions, they obviously do not provide an instantaneous solution to acute foreign-debt crises. Thus, other more immediate measures, such as debt rescheduling or new foreign loans, are usually sought. These measures are usually implemented with the help of the International Monetary Fund (IMF). Macroeconomic adjustments are normally conditions for getting such new loans and are often referred to as "conditionalities." The rationale for such conditionalities is that it makes little sense to make more loans available to provide a temporary solution to current foreign-exchange crises unless the causes of the crises are also dealt with.
The impact of economic crises and macroeconomic adjustment policies is likely to differ between population groups and thus affects income distribution whether intended or not. Yet there is virtually no systematic evidence on the distributional impact of these policies or of stabilization in general . Estimates of the impact on the poor and their nutritional status are also very scarce. Furthermore, because of lags in the economic relationships, the impact on a given population group as well as on the economy as a whole is likely to change over time; that is, short-term effects will differ from those in the longer term. Again, available empirical information on this matter is very scarce.
One of the reasons the impact on the poor and their nutritional status is poorly documented is that it is extremely difficult to estimate with a reasonable degree of accuracy. The relevant factors and relationships are many, and the separation of the impact of specific policy measures from the impact of other factors is very difficult.
The expected short-term impact of macroeconomic adjustment policies on the poor and their wellbeing, including their nutritional status, is usually not explicitly considered at the time of policy design, except as needed to maintain political stability, and does not appear to play a significant role in the design of conditionalities. Thus, in a review of IMF activities, Killick concluded, "Strengthening of the balance of payment situation was invariably the primary aim. Objectives concerning inflation and growth were secondary, while the distribution of income and wealth scarcely featured at all among the stated aims" .
Expected long-term effects on the economy as a whole play an important role in deliberations leading to the design and implementation of adjustment policies. Long-term effects on the various population groups, if explicitly considered, are usually assessed on the basis of capital investment-led growth models. These models imply that short-term cuts in consumption are necessary to assure improved consumption levels in the future and that shifts in income distribution in favour of the higher income groups will increase savings and thus productive investment. This argument is presented very clearly by Tseng , who argues that "the costs of adjustment, therefore, must be seen in perspective. In the short term, the costs are the unavoidable sacrifices that accompany the correction of an unsustainable situation." Even if the ultimate benefits to the poor of adjustment programmes could more than offset their short-term losses, however, the absolute poor may be unable to deal with such losses, even with the expectations of large longer-term gains, because they are already operating at the minimum subsistence level with few or no opportunities for borrowing to carry them over until the long-term benefits materialize. This is why the short-term effects on the poor are so important.
The need for capital accumulation also implies that rapid increases in economic growth will initially worsen the relative distribution of incomes. Thus, it should be no surprise that the poor run a high risk of short-term losses from adjustment policies aimed at improving the balance of payments and stimulating economic growth unless something is done to avoid it.
As mentioned above, short-term effects on specific population groups, such as the poor, are usually ignored unless they are perceived to threaten political stability. Yet, because adjustments frequently include changes of particular concern to the poor (e.g., increasing food prices and declining government expenditures on social programmes), those effects can be severe. Dis-investment in human capital through deterioration in nutrition, health, and education in the short run will hamper future economic growth and thus counter long-term solutions. Investment in human capital must accompany investment in physical capital. Finally, policies that cause severe hardships on the poor may result in political instability, which, in turn, reduces the impact of economic growth.
CHARACTERISTICS OF MACROECONOMIC ADJUSTMENT POLICIES
Macroeconomic adjustment policies are of five general types: monetary, fiscal, exchange rate, foreign trade, and wage and price.
Monetary policies are usually focused on limiting the expansion of the money supply and domestic credit availability, and increasing savings. Measures to influence interest rates and restrict credit for certain sectors (often the public sector) and activities are part of this group of policies.
Fiscal policies are usually aimed at reducing deficit government spending through restraints on expenditures and, to a lesser extent, through expanded government revenues. Measures to reduce expenditures include decreases in explicit subsidies on food and other goods and services; cuts in social programmes such as health care, education, and direct income transfers; and efforts to reduce deficits in public-sector enterprises. Reduced growth or reductions in public-sector employment and wages are also frequently found in adjustment programmes. On the revenue side, some adjustment programmes include measures to expand tax revenues and/or reform the tax structure.
Adjustments in the exchange rate are common in adjustment programmes. At the time when macroeconomic adjustments are needed, the national currency is usually overvalued and devaluations are called for.
Several measures aimed at foreign-trade reforms are found as components of adjustment programmes. These include changes in export taxes and import duties, export promotion, etc. Finally, adjustment programmes usually include measures to restrain wage increases and to deregulate prices.
The nature of the policy measures is discussed further in the next section.
CONCEPTUAL OVERVIEW AND REVIEW OF AVAILABLE EVIDENCE
A conceptual overview of the principal relationships between human nutrition and variables influenced by economic crises and macroeconomic adjustment policies is presented in figure 1.
The nutritional status of an individual is influenced by his or her health and food consumption. Health, in turn, is influenced by the availability of health, sanitation, and related programmes. The extent to which such programmes are available depends on the amount of government expenditures allocated to them. During periods of economic crises, which are often associated with or caused by excess government spending, governments frequently try to cut spending on social programmes, including those for health care, as part of macroeconomic adjustment policies.
FIG. 1. Schematic overview of principal relationships between human nutrition and variables influenced by economic crises and macroeconomic adjustment policies
Government expenditures on health care in the developing countries as a whole decreased from about 4.3 to 4.1 per cent of total government expenditures during 1976-1983 [3, 4]. The decrease was most severe in Latin America. No data are available for Africa. The average figures cover large differences between countries. In Argentina and Mexico, government health expenditures dropped from more than 4 per cent to a little more than 1 per cent over a seven-year period. The real value of government expenditure on health per capita also decreased dramatically in these countries [3, 4].
Several other countries with severe economic crises and subsequent macroeconomic adjustments show similar decreases. Thus, in Jamaica the real value of health expenditures fell by 35 per cent from 1982 to 1985 . These cuts may have been a direct consequence of severe macroeconomic adjustments made during that period. Data on average real values of government expenditures in health per capita for Latin America, Africa, Asia, and developing countries as a whole were not readily available, but it is likely that there has been a slight increase in Asia and a decrease in Latin America.
Large cuts in health-care expenditures would be expected to have serious effects on the health and well-being of the poor unless they either occur primarily in the services used by the better-off population groups or are accompanied by improvements in programme efficiency. The distribution of the cuts among types of health services varies among countries, and available evidence does not provide general guidelines regarding the extent to which the poor are affected.
In several countries, including Chile and Jamaica, investments in the health sector were reduced considerably more than recurrent costs. It is possible that investments are biased toward the better-off population groups because they tend to use more capital-intensive health care, such as hospitals and special equipment, while recurrent costs may be biased toward the poor, e.g., primary health care. If this is correct, the poor may have been at least partially protected. Jolly and Cornia  concluded that there was no evidence in the countries studied by UNICEF of "restructuring of public spending toward cost-effective, high-priority action which can protect children even in times of recession and cutbacks."
In Brazil, there was a major shift in relative expenditures on curative versus preventive health care during the 1970s and 1980s . The latter decreased from about 40 per cent of total health expenditures in 1970 to about 15 per cent in 1982. A similar although less rapid relative shift took place during 1978-1982 in the Philippines, where the percentage of total health-care expenditures used for field health services stayed constant at about 28 per cent while expenditures on hospitals increased from 48 to 54 per cent . The cost of general administration and planning appears to have decreased.
When preventive health care is primarily focused on the poor while expenditures on curative health care are biased toward the better-off population groups, the above relative changes appear to have had adverse effects on the poor. Available evidence is insufficient to provide reliable estimates of the effects of reduced government expenditures on the poor and their nutritional status.
In many developing countries, a large share of total healthcare expenditures originates in the private sector either as employee benefits or from households directly or through insurance schemes. The impact of macroeconomic adjustments on employee benefits is discussed together with wages in the next section. Direct household expenditures on health care are typically 3 to 5 per cent of total household expenditures in developing countries. The price elasticity of demand for health care appears to be low . Thus, the introduction of fees for health-care services in order to reduce government spending is likely to diminish the use of such services relatively little but instead to increase total household expenditures on health care. This, in turn, reduces the purchasing power available for other goods and services such as food.
Food consumption by individuals at risk of being malnourished is determined by household food consumption and intra-household food allocation. The intra-household food allocation in turn is influenced by a number of factors. Among those that may be affected by macroeconomic adjustment policies, the most important are probably time allocation, employment, wages, and income control for each of the adult household members.
Macroeconomic adjustment policies may cause changes in employment opportunities and wages for women. Such changes may in turn alter the amount of time spent on child care, meal preparation, and other related home activities. They may also alter the distribution of income control between men and women. If, as indicated by some studies, spending behaviour differs between men and women, changes in the distribution of budget control may affect the proportion of total household incomes spent on food as well as the amount of food allocated to individuals, for example, pre-schoolers at risk of malnutrition. Changes in women's employment opportunities, and thus demands on their time and the opportunity cost* of their time as well as changes in their control over household expenditures, may alter diet composition and cost per calorie or nutrients. For example, increased employment opportunities and/or higher wages for women tend to reduce the time they have available to prepare meals and increase the calorie cost of the diet because more semi-processed or processed foods, such as wheat bread instead of sorghum and millet in certain countries, are likely to be purchased. On the other hand, more is likely to be spent on food both because of higher household incomes and because a larger proportion of income controlled by women may be spent on food. Increasing employment opportunities for women may also reduce the opportunities for breast-feeding because nursing facilities are often lacking at the workplace.
The key factors influencing household food consumption are household real incomes and the prices of food and non-food commodities the household wishes to acquire. Low-income consumers spend a large share of their incomes on food; 60 to 80 per cent is not uncommon. Furthermore, the income elasticity of demand for food - that is, the percentage change in food demand for a 1 per cent change in income - is much higher for the poor than for the better-off. Therefore, food consumption by the poor is particularly hard hit by decreasing incomes.
Editor's note: When time spent on one activity precludes another activity, the value of the two activities can be compared. Thus, for better or worse, one activity costs the opportunity to engage in another simultaneously. If one cannot have both A and B. the choice of A excludes the choice of B.
To assess the impact of economic crises and macroeconomic adjustment policies, it may be useful to disaggregate household incomes not only by gender as discussed above but also by type and source of income, distinguishing:
- monetary income, including wage earnings, incomes of the self-employed, and gifts and transfers, and
- in-kind income, including consumption of one's own production, in-kind wages, and in-kind gifts and transfers.
Economic crises and adjustment policies may affect any of these types of income, but the effect will differ among policy measures. Furthermore, the effect of a certain income change on food consumption and nutrition is likely to depend on whether the change occurs in wage earnings or in some other income source.
Wage earnings depend on the number of hours or days worked, that is, the amount of employment, and the wage level. Most macroeconomic adjustment programmes attempt to expand employment while keeping wages from increasing. Wage and price policies are often focused on reducing real wages (i.e., the nominal wage relative to the price level) in order to expand employment and boost savings in business and among high-income groups. Wage controls are also enforced to restrain the inflationary pressures generated by other components of the macroeconomic adjustment programme, particularly currency devaluations.
If successful, these policies will bias the income distribution toward the better-off, at least in the short run. The impact on poverty will depend on the relative magnitudes of employment increases and wage declines for the poor. In Chile, both relative income distribution and absolute poverty deteriorated between 1969 and 1978, a period of serious economic crisis and severe macroeconomic adjustment. Thus, the real value of total expenditures of the poorest quintile of the population dropped by 30 per cent, while for the richest quintile it increased by 16 per cent (fig. 2). From the point of view of equity as well as that of poverty and nutrition, such developments should be of great concern, and efforts should be made to determine the causes and design corrective and/or compensatory measures.
Real wages have decreased in many countries after macroeconomic adjustment policies were introduced. Thus, real wages fell in Sri Lanka during the period 1979-1984 after adjustments were made in 1977-1978 [8, 9]. During 1981-1984 real wages declined in five of seven Latin American countries studied by the Inter-American Development Bank . The decline was most dramatic in Mexico and Brazil, where the cumulative decline over the four year period exceeded 20 per cent (table 1). The real wage decreases were not offset by increasing employment. On the contrary, unemployment increased in most of these countries (table 1). Thus, wage earnings declined during the period.
FIG. 2. Change in total monthly mean expenditure per household (in December 1979 pesos) by income quintile in Chile, 1969-1978 (Source: Jolly and Cornia )
The extent to which the poor- as opposed to better-off wage labour- were affected is not clear from available evidence. Drastic decreases in the real minimum wage rate in a number of countries indicate, however, that the poor are seriously affected. Thus, the purchasing power of the minimum wage in Zaire in 1982 was only about 3 per cent of what it was in 1970 . Furthermore, the Inter-American Development Bank concluded that there is evidence that a disproportionate part of losses in real incomes "have been concentrated in the lower income strata." The Bank further suggested that "to the extent that real wage containment remains a necessary element of the adjustment process, mechanisms will have to be found to shift some of the burden to the higher income groups in the interest of social justice and domestic peace" .
In many developing countries, the poorest wage labour is found in the informal sector producing goods and services that are not exported or imported- the so-called non-tradables. Since macroeconomic adjustment policies aim at the improvement of the balance of payment, they typically increase the prices of goods and services that are either exported or imported- the tradables- relative to the non-tradables. Therefore, if the poor are concentrated in the non-tradables sector, they are likely to be made worse off relative to wage earners in the tradables sector. If these poor workers are able to move between sectors and thus take advantage of relatively higher wages and better employment opportunities in the production of tradables, the problem may be less severe. Thus, policies to increase the mobility of poor workers are of great importance.
As mentioned above, a large share of the poor are self-employed, such as small farmers and small entrepreneurs, the latter mostly in the so-called informal sector. As the prices of tradables increase, those self-employed who produce for export or for import substitution, such as small farmers producing export crops, are likely to benefit from changes in the exchange rate and related adjustment policies, for example, export-promoting policies. Furthermore, those engaged in the production of inputs for the tradables sector may gain. The most likely losers are those who produce final goods and services that are neither exported nor imported, because their prices are unlikely to increase as fast as the prices of other commodities. Therefore, the real incomes of those population groups decrease. Unfortunately, although statistical information is scarce, it appears that a large proportion of the poor are occupied in precisely such production.
Gifts and transfers are another important source of income for the poor. Government transfer programmes take many forms, including explicit price subsidies for selected commodities such as food, housing, and transportation; food stamps; unemployment compensation; social security payments; and poverty relief payments. The importance of these transfers is illustrated by food subsidies, which account for 15 to 20 per cent of total income of the poor in a number of countries 112]. Macroeconomic adjustments typically attempt to reduce subsidies and other transfers in order to reduce government expenditures and remove price distortions.
TABLE 1. Changes in real wages and urban unemployment in seven Latin American countries, 1981-1984
|Real wages (% change from previous year)||Urban unemployment (% of labour force)|
|Argentina||- 10.3||- 10.5||19.1||5.3||4.5||4.7||4.0||3.8|
|Mexico||4.9||- 4.1||- 20.0||- 9.8||4.2||4.1||6.9||6.3|
For developing countries as a whole, the percentage of government expenditures allocated to transfer programmes stayed constant during the period 1976- 1982 .
Some countries show drastic decreases in the real value of transfer programmes. Thus, the Sri Lankan expenditures on such programmes fell from 20 per cent of total government expenditures in 1976 to 12 per cent in 1982 [3, 4]. Furthermore, the value of the food-subsidy programme fell from 16 per cent of current government expenditures in 1977 to 3 per cent in 1984 .
By 1982 the real per capita value of transfer programmes in Sri Lanka had been reduced to about half of the real value in 1978, a dramatic change indeed, which is almost certain to have caused severe hardships among the poor. The drop was even more severe in Ghana [3, 4]. On the other hand, expenditures on transfer programmes increased significantly in Argentina during the same period.
In-kind incomes are an important source of income for many poor households, particularly semi-subsistence farmers. The importance becomes very clear during periods of rapid increases in the prices of food and other necessities that frequently follow macroeconomic adjustments. Price increases erode the real value of monetary incomes such as money wages. Thus, as food prices in crease, a given wage income buys less and less food. Such erosion does not occur for a given in-kind income.
Developments in Uganda during the last 10 to 15 years illustrate this point . The price index rose from 100 in 1972 to 35,000 toward the end of 1984. The price of maize meal, the principal food staple, rose from about 1 Ugandan shilling per kilogram in 1972 to 180 shillings per kilogram in early 1985. The rate of increase in wages was considerably less. Thus, the index for real wages fell from 100 in 1972 to 9 in late 1984. The number of minimum wages needed to purchase the food necessary to meet minimum calorie requirements for an average household increased from 0.6 to 4.5 during the 12-year period.
In spite of these developments, however, the nutritional situation did not deteriorate greatly during the period. The explanation as provided by Jamal is that most of the rural poor obtain a major share of their food from their own production and the urban poor greatly increased food production for their own consumption.
These findings clearly illustrate the nutritional risk associated with the shift from subsistence to a cash economy. While such a shift is a necessary condition for significant improvements in the standard of living, it is important that the nutritional risks caused by increased dependence on prices and monetary incomes should be countered through appropriate policies. While macroeconomic adjustment policies usually are not explicitly concerned about the above shift, they typically promote increases in the production of export crops, sometimes at the expense of food production for one's own consumption. Thus, depending on the relative price changes in food staples and export crops, the nutritional risk may change . (The nutritional implications of increasing commercialization of semi-subsistence agriculture are currently being studied by the International Food Policy Research Institute [IFPRI] in several countries.)
In-kind incomes may also play a major role among agricultural workers, who often receive part of their wages in food and housing. To the extent that these in-kind wages are kept unchanged during periods of rapidly increasing prices, the nutritional status of the workers and their families is protected, at least in part, from the negative effects of the price increases. There are indications, however, that the share of farm wages paid in kind falls in periods of rapidly increasing food prices and/or decreases in production. This point is illustrated by findings from the North Arcot region of Tamil Nadu, India, where landless workers received 11 per cent of their total household calorie consumption from in-kind wages (primarily rice) in 1983184, which was a year of normal production, as compared to less than 1 per cent during the year before, a drought year .
It is also important whether government transfers are in kind or in money. Food subsidies in Sri Lanka illustrate this point. Until 1978, fixed amounts of rice were made available to households either free or at a subsidized price. Thus, irrespective of price fluctuations, households were assured a certain amount of food. In 1979 the programme was changed to a food-stamp programme with a fixed monetary value. Since then, food prices have quadrupled while the monetary value of the food stamps received by the household has remained constant. As a consequence, the amount of food that can be purchased by the food stamps today is only about 25 per cent of the amount that could be bought in 1979.
As mentioned above, real incomes are determined by nominal incomes and prices. Two aspects of prices are important in this regard: (1) the level of prices and changes in the level, usually referred to as inflation, and (2) the relative prices of various goods and services.
Economic crises and adjustments may influence price levels in a variety of ways, most importantly through the demand and supply of goods and services, the supply of money, and the exchange rate. Although the situation may vary among countries and time periods, one grossly simplified example may be as follows: Government deficit spending financed by expanded domestic money supply and external borrowing along with (1) depressed export prices and/or sharp increases in import prices, (2) structural problems in the production sectors, and (3) investment patterns with very limited short-term returns cause or contribute to a high rate of inflation, increasing balance-of-payment problems, and an overvalued exchange rate. As the balance-of-payment problems become more severe, the country attempts to make the necessary macroeconomic adjustments, including devaluation, reduced government spending, export promotion, structural adjustments in the production sectors, increased domestic interest rates, wage freezes, and possibly other policy measures. The net impact of the crisis and the adjustments on prices is not easily determined. Expanded money supply, increased external borrowing, and devaluation all tend to cause inflation. Thus, key components of the economic crisis as well as a key component of adjustment efforts would be expected to cause increasing price levels. Other components of the adjustment programme, such as structural improvements in the production sectors, higher interest rates, and reduced money supply, are likely to dampen inflation.
Increases in the general price level have been dramatic in many developing countries during the recent past, as illustrated by data from Brazil, the Philippines, and Zaire (table 2). Thus, the consumer price index in Brazil increased from 100 to 60,000 during the period 1970-1984. Food prices increased even more, from 100 to 88,000. While the impact on the purchasing power and thus the ability to acquire sufficient food to meet nutritional requirements is clearly negative, available empirical evidence does not provide reliable information on the relative contribution of the various policy measures to this high rate of inflation. Without such information, it is difficult to predict the nutritional implications of specific macroeconomic adjustment policies. In fact, without trying to estimate causal effects, it is difficult to establish whether the economic crisis or the attempted solutions had the most negative nutrition effects.
This question may be only of academic interest, however, because a continuation of the economic crisis without adjustment might not be viable. Rather, the question may be one of the impact of various policy options.
TABLE 2. Consumer price indexes for all commodities and for food in Brazil, the Philippines, and Zaire, 1970-1984 (1970 = 100)
|Consumer price index||Consumer price index for food|
Sources Refs. 11, 16, and 17.
NA = not available.
TABLE 3. Impact of a 10 percent increase in the price of food on real incomes of low- and high-income population groups
|% decrease in real income||Sourcea|
|Lowest docile||Highest docile|
|Sri Lanka||8.5||4.1||Sahn |
|Egypt||5.6||1.0||Alderman and von Braun |
|India||7 3||2.9||Murty |
|Funtua, Nigeria||7.7||6.5||Pinstrup-Andersen and Uy |
|Gusau, Nigeria||9.0||5.7||Pinstrup-Andersen and Uy |
Source: Ref. 24.
a. Sources of data used for calculations, except for Mellor, who reported the estimates shown.
b. Food grains only
c. For the lowest 20 per cent.
d. For the highest 5 per cent.
The negative impact of food price increases is much more severe for the poor than for the better-off because the poor spend a larger budget share on food. Furthermore, the consequences of reduced food consumption among population groups consuming near the subsistence level may be further deterioration in nutritional status. The more severe impact on the poor is illustrated in table 3. For example, in Egypt, a 10 per cent in crease in the price of food causes a 5.6 per cent reduction in real incomes of the poorest decile but only a 1 per cent reduction in real incomes of the richest docile.
The greater impact of food price changes on the nutritional status of the poor is also illustrated by the larger price elasticities of demand for food, that is, the percentage change in food demand associated with a 1 per cent change in food prices. As shown in table 4, increases in the price of rice cause much larger percentage reductions in rice demand among the poor than among the rich.
Partly because of the importance of food prices in the consumer budgets of the poor, many governments have attempted to keep these prices low through implicit or explicit price subsidies. Implicit price subsidies, however, which are usually enforced through low producer prices, have adverse effects on food production, while explicit subsidies, usually financed by government, may contribute to excess government spending. The latter contributes to inflation, and the net effect of explicit food price subsidies may be to increase rather than decrease prices . Of course, households that spend a large budget share on the subsidized commodities may gain from the subsidy in spite of an increase in the general price level. Thus, the choice of commodities to be subsidized as well as targeting on the poor are important considerations.
TABLE 4. Price elasticity of demand for rice among low- and high-income population groups in selected countries
|Low income||High income||Source|
|Percentile||Price elasticity||Percentile||Price elasticity|
|Colombia (Cali)||1||-0.43||93||-0.19||Pinstrup-Andersen, et al. |
|Indonesia||8||- 1.92||55||-0.72||Timmer and Alderman |
|Bangladesh (rural)||10||- 1.30||90||-0.83||Ahmed |
|Thailand||12||- 0.74||87||- 0.46||Trairatvorakul |
|Brazil||15||- 4.31||90||- 1.15||Williamson-Gray |
|India (rural)||3||- 1.39||96||-0.39||Murty |
|India (urban)||1||-1.23||92||-0.21||Murty |
Source: Ref. 24.
Contents - Previous - Next