Contents - Previous - Next

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

Economic policies and the nutritional status of the urban poor in Chile, 1968-1976

Oscar Ruiz

The food buying power index (ICCA: el indice de capacidad de compra de alimentos)
The food buying capacity index (ICCA)
Economic policies and their effect on the food buying capacity of the urban poor
The food buying capacity index and trends in the nutritional situation of the urban poor in Chile


In Chile in 1968, 47 per cent of the urban and 78 per cent of the rural population had incomes lower than US$100 a month and were considered poor. At the same time, a tiny minority comprised of 3 per cent of the urban and 1 per cent of the rural families enjoyed monthly incomes that were ten times greater (1). If the population characterized as poor is considered as a whole, it represented five million people, or approximately 56 per cent of the total population (see Table 1 in the Statistical Appendix to this chapter, p.146).

If a level of US$50 monthly income is chosen as the limit for absolute poverty, about 36 per cent of the poor urban and 57 per cent of the rural families fell within this range-approximately 2,272,000 people, or about 25 per cent of the total population of the country. What effects have diverse economic policies applied between 1968- 1976 had on food availability and nutritional status of this poor segment of the population? How does one choose a satisfactory method for analyzing these effects using an economic and nutritional approach? The study reported here was undertaken to answer these questions. It is divided into four parts.

The first part describes the search for an adequate method to analyze the trends in the nutritional status of the urban poor, incorporating combined economic variables (income and prices).

The second part analyzes the various economic policies applied during this period, and their effects on the food buying power of the urban poor.

The third section of the paper includes an analysis of the Food Buying Power Index to vailidate its interpretation of trends in the nutritional status of the poor.

Finally, the fourth and final part draws the most important conclusions that can be derived from the first three sections.

The analysis in the first part includes the results of research conducted by the author in 1978 at the Institute of Nutrition and Food Technology of the University of Chile in collaboration with the United Nations University (2).

The food buying power index (ICCA: el indice de capacidad de compra de alimentos)

The main interest behind this study was to measure the effects of economic variables relating to food prices and personal income on the nutritional status of the poor.

The lack of continuous, reliable information in the rural sector limited the analysis to the poor in urban sectors, for which relatively abundant information existed. Nevertheless, the lack of appropriate indicators for an analysis of the variables pertaining to nutritional conditions among the poor in urban areas required the construction of an indicator for assessing food buying power. This variable had to be easily connected to the nutritional status of this part of the population. This indicator, called the Food Buying Power Index (ICCA), was conceived in order to compare tendencies in the evolution of income and of food expenditures of the urban poor, period by period

For development of this index, two indices had to be selected. One was a personal income index, the other an index of expenditures for food. They each had to reflect adequately the variations in amount of income spent for food by the urban poor. These two indices combined with the evolution of the ICCA are now examined separately for the 36 quarterly periods between 1968 and 1976.

A. The Selection of the Income Index

The amount of reliable information available only permitted consideration of the series of indices that measure the variations in personal income within two categories of workers: white collar (office employees) and blue collar (labourers). These correspond to the series developed for quarterly use by the National Institute of Statistics (INE)* in 1960 that have been employed ever since for calculating the wage indices and salary indexes for blue- and whitecollar workers (3).

The basic methodology of both calculations corresponds to the quarterly measurement of the variations in average incomes paid by the companies to their labourers and employees. For this purpose, in 1960 the INK carried out a sampling of 421 industrial until 1970, the INK was called the Department of Statistics and Census. establishments with over 20 workers each, amounting to a total of 228,527 labourers and 106,467 employees.

The preliminary analysis of the patterns noted in both series between 1968 and 1976 indicated different trends in some periods. The next problem was to choose a method that would best represent the variations in the income of the urban poor. For this purpose, the average income level represented by both indices, and the characterization of poor homes by occupation of the head of the household, were analyzed. This analysis concluded that in poor homes, heads of household in the labourer category corresponded to 41 per cent of the poor homes, and were more than double the number of households with an office employee as the head. These comprised only 18 per cent of the total number of poor homes.

Furthermore, the analysis of average incomes showed that the average income for a labourer's household was similar to that for poor white-collar workers. Tables 2 and 3 in the Appendix show these figures in more detail. Income analysis determined, in the end, the adoption of the wage index of the INK.

The selection of the series developed between 1968 and 1976 for this index was not without difficulties. These problems related to whether they truly reflect actual income variations in the remaining categories of workers and among the unemployed. For these, statistical information did not exist at the time this analysis was undertaken. However, when the series of blue-collar wage and white-collar salary indices are compared, the former seem to be more representative, for reasons already stated, of the trends in income variation among the urban poor

B. The Elaboration of a Monthly Food Spending Index (IGA: Indice de Gasto Mensual en Alimentos)

A group of food items had to be chosen in order to characterize the food expenses of the urban poor. The quantities of these were fixed so that the variations in total spending caused by price fluctuations in each period could be measured. All of this was possible through selection of a food basket following certain economic criteria within a reference frame already established using nutritional criteria.

This task was undertaken in three steps. The first consisted of the adoption of nutritional criteria that could lead to the determination of a ration per group of foods in grams per day for an average man. The total population indicated in the 1970 census by age, sex, physiological status, and activity level (4) was divided into 28 nutritional categories. To each of these, a model ration {which had previously been established) (5, 6) was assigned. These rations were to satisfy nutritional needs by distributing the consumption among 12 groups of foods. The rations included information on the consumption habits of the population.

Once this distribution was made (Table 4, Appendix), the rations were weighted with the percentage of the total population associated with each model ration. This allowed us to rely on a ration with recommended daily allowances of 12 groups of foods for the average Chilean population, as follows.

TABLE A. Model Ration for an Average Person per Food Group

Food group Grams
Milk 520
Meat 53
Fish 38
Eggs 24
Legumes 28
Potatoes 223
Green vegetables 163
Fruit 175
Bread 301
Flour 41
Sugar 38
Fat 35

TABLE B. Model Ration for an Average Person (AP), per Nutrient Type

  AP per day
Calories 2,215
Proteins (9) 65
Lipids (9) 69
Carbohydrates (9) 355
Calcium (mg) 585
Iron (mg) 12.5
Vitamin A (IU) 4,300
Vitamin C (mg) 60
Thiamine (mg) 7.10
Riboflavin (mg) 1.60
Niacin (mg) 15.30

This ration, in turn, satisfies the requirements for the nutrients shown in table B, in the units indicated, per day.

After this task had been completed, it was necessary to take a second step, which consisted of assigning the recommendations in grams per day (provided by each of the 12 groups of the average ration) for the corresponding specific foods.

In order to determine which foods, and in what quantities, should be included in each group, it was necessary first to analyze the spending structure and the categories of elasticities shown by the various income levels. This was possible through the utilization of information on budgets and urban family expenses.

One study (7) showed that 98 per cent of money spent on food among the population with incomes (in 1968) of less than US$100 per month was destined to increase the existing consumption levels for all items under study. This illustrated clearly the extremely low capacity of the poor to diversify this kind of expense. It also revealed the underlying existence of gross under-nourishment.

If a level of income between US$100 and US$200 per month is analyzed, an expense structure totally different from the above one is perceived. In this structure, 50 per cent of the increases in spending rose at moderate rates (elasticities between 0 and 1), and 20 per cent of total spending rose at rates more than proportional (elasticities greater than 1). This was compensated for by some 21 per cent of money spent on food, which reduced the demand for the foods being considered (with negative elasticities.-Table 5, Appendix.)

If the average calorie and protein consumption levels of the income bracket below US$100 are considered simultaneously and compared to the average recommendations, malnutrition is again detected. To this should be added the inequities in intra-family food distribution according to income (Table 12, Appendix).

The need to be able to rely on a diversified spending structure that would not be biased by malnutrition and the relative scarcity of most of the food items considered, and also the great disparity that exists between average consumption of families at an income level lower than US$100 and the recommended daily allowances, determined the choice of a level of spending between US$100 and US$200 as the most convenient to represent the food expenditures of the poor sectors. Even though this doubled the standard for poverty level, it allowed for easier establishment of priority criteria for selecting, from a large group of foods, those that must be included in a basic basket having a limited number of items. This could not have been achieved by considering a level of income below US$100 because the group of foods was even more limited, and furthermore, there was a relative scarcity of these items, as shown by the spending structure at this level.

Therefore, by considering the spending structure indicated in the Study of Budget and Expenses per Family of 1968 (8), we proceeded to select a group of 37 food items that were periodically registered, and that were not subject to strong seasonal variations in demand.

The quantities assigned to each food were computed by the division of the average spent on each item by its price. This created a common base for each food group, expressed in grams. According to their contribution to the total (in percentages), each food item was assigned a corresponding share in grams of the recommended intake for each group. This basic basket was then calculated for a 30-day period, and the quantity of each selected item was expressed in the respective retail buying unit.

Once the foods were selected and the respective quantities were determined, the calculation of an index of expenses could be made. For this purpose, an index of the Laspeyres type (Appendix, Item VIII, Note 1) was used. In this case, the variations in the index are the result of fluctuations in total expenses produced by keeping the quantities (the ration) at a fixed level and by modifying the prices.

The basis for the index corresponded to total spending for the basket, or ration, in the first quarter of 1968. The corresponding weight of each food was calculated by the percentage assigned to each one with respect to total spending in the basic period.

The prices applied were those corresponding to the price average registered for each quarter by the INK. Table 6 in the Appendix shows the rations for each one of the 37 items and their corresponding weights. In order to create the expense index, we compared the variations in total spending in each period, and the total spending in the basic period.

The food buying capacity index (ICCA)

Once the expense index was made, we built up the series through the ICCA for each one of the 36 quarters between 1968 and 1976. As indicated before, this index initially measures the relationship between variations in income and food expenses of the urban poor. This is a means for quantifying the impact of economic variables, mainly income and food prices, on the nutritional status of the population.

The ICCA is constructed to show a simple relationship between the variations produced between two indices within one period: income and spending in the period analyzed with the same relationship to the reference period (Note 2, VlII, Appendix). Increases over 100 in the index indicate an improvement in food buying capacity; decreases indicate the opposite. The results of the ICCA for each one of the 36 quarters are as follows.

TABLE C. Evolution of the Food Buying Capacity Index (ICCA) for a Poor Urban Population between 1968 and 1976

Year Quarter Index of wages Index of expenses ICCA2
    1st quarter 1st quarter 1st quarter
    1968 = 100 1968 = 100 1968 = 100
1968 I 100 100 100
  II 106 105 102
  III 115 102 105
  IV 123 114 107
1969 I 140 126 111
  II 160 142 113
  III 161 145 111
  IV 166 158 105
1970 I 187 191 97
  II 207 189 110
  III 221 205 109
  IV 251 229 110
1971 I 273 232 117
  II 316 236 133
  III 350 256 137
  IV 374 280 133
1972 I 426 324 131
  II 476 374 127
  III 523 453 115
  IV 841 1,145 73
1973 I 1,200 1,550 77
  II 1,375 1,517 90
  III 1,988 2,684 73
  IV 2,543 6,774 38
1974 I 6,490 9,383 69
  II 8,368 13,227 63
  III 14,453 21,433 67
  IV 19,120 34,134 55
1975 I 26,338 47,083 56
  II 40,933 74,266 55
  III 64,648 114,359 56
  IV 85,040 168,406 50
1976 I 112,075 216,630 52
  II 157,361 297,201 53
  III 225,727 406,793 55
  IV 297,641 501,524 59

1 Index of wages furnished by INK
2 Index of expenses and ICCA according to the methodology indicated in notes 1 and 2 in Appendix.
(See also Figure 1.)

Contents - Previous - Next