Cover Image
close this bookOvercoming Child Malnutrition in Developing Countries - Past Achievements and Future Choices. 2020 Vision for Food, Agriculture, and the Environment. Discussion Paper 30 (IFPRI, 2000, 73 p.)
View the document(introduction...)
View the documentForeword
View the documentAcknowledgments
View the document1. Exploring the Causes of Malnutrition
View the document2. Determinants of the Nutritional Status of Children
View the document3. Data and Methods
View the document4. New Evidence from Cross-Country Data, 1970-95
View the document5. How Has Child Malnutrition Been Reduced in the Past?: A Retrospective
View the document6. Projections of Child Malnutrition in the Year 2020
View the document7. Priorities for the Future
View the document8. Conclusions
View the documentAppendix: Cross-Country Studies: Methodological Issues and Past Findings
View the documentReferences
View the documentRecent Food, Agriculture and the Environment Discussion Papers

7. Priorities for the Future

Chapters 4 and 5 examined the past record of reductions in child underweight rates and attempted to isolate the contributions of the four variables representing its underlying determinants and two variables representing its basic determinants. Chapter 6 then developed three scenarios for child malnutrition in the year 2020, based on past trends in the underlying-determinant variables. The scenarios are essentially the answer to the question “If we continue as in the past (or perhaps do a little more or a little less), what will the future look like?”. Even under the most optimistic of the scenarios, the prevalence of child malnutrition in developing countries is expected to be 15 percent in 2020: 128 million children would still be malnourished. But the future doesn’t have to look like the past.

This chapter asks, for each developing region: “What combinations of actions would lead to the greatest reductions in child malnutrition by 2020?” In answering this question, it is important to keep three things in mind. First, as the conceptual framework of this paper lays out and the analysis has confirmed, all three underlying determinants - food security, mother and child care, and a healthy environment - are necessary for a child to achieve adequate nutritional status. Thus strategies for reducing child malnutrition should address all of them. The issue taken up here is the relative emphasis that should be placed on the various contributing factors.

Second, both underlying and basic determinants were found to have strong effects on child malnutrition, with the former being dependent on the latter. The question is not which set of determinants should be prioritized: both underlying and basic determinants should be the focus of future efforts to reduce child malnutrition.

Finally, actions associated with the determinants considered in this report should be seen as complementary to the more direct measures that have been the traditional focus of nutrition interventions, such as promotion of breast-feeding, nutrition education, supplementary feeding, and food fortification (see Gillespie, Mason, and Martorell 1996). In addition, the reader should keep in mind that increased per capita dietary energy supplies and per capita national incomes can be brought about not only by raising food supplies and national incomes, but also by reducing population growth.

Relative Importance of the Underlying-Determinant Variables to Future Reductions in Child Malnutrition

Table 14 compares, for each region, the strengths and potential impacts of the variables chosen to represent the underlying determinants of child malnutrition. In column (2), calculations of the absolute increase in each determinant needed to bring about a reduction in the child malnutrition rate of one percentage point in 1995 are presented, 20 For example, in South Asia an increase in the rate of access to safe water of 13.1 percentage points would have the same effect on child malnutrition rates as would an increase in per capita dietary energy supply of 94 kilocalories. The different range of values each determinant takes on makes this column difficult to interpret when comparing across determinants. Therefore, the absolute increases in column (2) are standardized by reporting them as a percent of the determinants’ ranges. This number, given in column (3), is the measure of strength of impact.

20 The 1995 numbers differ from those presented in Table 8 only for the variable per capita dietary energy supplies. This variable’s strength of impact depends on its current level (see Table 6).

Table 14 - Comparison of the strengths and potential impacts of factors affecting child malnutrition, 1995

Region/variable

1995 mean
(1)

Increase in variable needed to reduce prevalence of child malnutrition by percentage point
(2)

Number in (2) as a percent of developing-country range a
(3)

Percent determinant is below its desirable level b (0-100 scale)
(4)

Change in prevalence of child malnutrition with increase in determinant to desirable level c
(5)




(percent)

(percent)

(percentage points)

Underlying-determinant variables


South Asia



Access to safe water (SAFEW) (percent)

79.7

13.1

13.2

-20.3

-1.6



Female secondary school enrollment (FEMSED) (percent)

34.1

4.6

4.6

-65.9

-14.5



Female-to-male life expectancy ratio (LFEXPRAT)

1.023

0.0139

9.3

-58.9

-5.5



Per capita dietary energy supply (DES) (kilocalories)

2,356

94

4.5

-46.5

-3.0


Sub-Saharan Africa



Access to safe water (SAFEW) (percent)

48.8

13.1

13.2

-51.2

-3.9



Female secondary school enrollment (FEMSED) (percent)

19

4.6

4.6

-81.0

-17.8



Female-to-male life expectancy ratio (LFEXPRAT)

1.054

0.0139

9.3

-35.2

-3.3



Per capita dietary energy supply (DES) (kilocalories)

2,136

75

3.6

-60.2

-5.2


East Asia



Access to safe water (SAFEW) (percent)

66.5

13.1

13.2

-33.5

-2.6



Female secondary school enrollment (FEMSED) (percent)

59.8

4.6

4.6

-40.2

-8.8



Female-to-male life expectancy ratio (LFEXPRAT)

1.0514

0.0139

9.3

-37.4

-3.5



Per capita dietary energy supply (DES) (kilocalories)

2,720

188

9.0

-23.8

-2.1


Near East and North Africa



Access to safe water (SAFEW) (percent)

81.5

13.1

13.2

-18.5

-1.4



Female secondary school enrollment (FEMSED) (percent)

57.9

4.6

4.6

-42.1

-9.2



Female-to-male life expectancy ratio (LFEXPRAT)

1.044

0.0139

9.3

-42.8

-4.0



Per capita dietary energy supply (DES) (kilocalories)

3,172

333

16

+4.5

-0.2


Latin America and the Caribbean



Access to safe water (SAFEW) (percent)

77.3

13.1

13.2

-22.7

-1.7



Female secondary school enrollment (FEMSED) (percent)

56.5

4.6

4.6

-43.5

-9.6



Female-to-male life expectancy ratio (LFEXPRAT)

1.098

0.0139

9.3

-1.9

-0.18



Per capita dietary energy supply (DES) (kilocalories)

2,777

234

11.2

-20.2

-1.8

Basic-determinant variables d


Per capita GDP (GDP) ($ PPP)

2,121

202

9.7

-59.1

-18.5


Democracy (DEMOC)

2.71

0.79

11.5

-71.5

-5.5

Note: The table compares the relative strengths of the underlying-determinant variables to one another and those of the basic-determinant variables to one another. Since the two groups lie at different levels of causality, it is important not to compare the results for variables across the groups.

a See Table 8 for variable ranges.

b The desirable levels of the variables are: SAFEW: 100 percent; FEMSED: 100 percent; LFEXPRAT: 1.1 (this Is the average of the highest 20 percent of country-year data points in the sample in terms of female-to-male life expectancy ratios [excluding the highest, which is 1.15 and far above the next highest, 1.12]); DES: 3,100 (see text footnote 21 for rationale); GDP: The desirable level is set at $4,750. This is the level past which improvements in GDP per capita no longer contribute to reductions in child malnutrition; DEMOC: 7 (the maximum value of the index).

c These numbers are calculated using the regression coefficients in Table 8 columns (3) and (4). For DES and GDP, each region’s number is calculated using country averages.

d Because the structural relationship between child malnutrition and basic determinants differs by regions, reliable regional breakdowns cannot be provided.

In South Asia and Sub-Saharan Africa, food availability (shown as DES) emerges as the determinant that needs to change the least, relative to its existing range, to bring about a 1 percentage-point drop in child malnutrition rates. It is thus the most potent force in reducing child malnutrition. Female secondary school enrollments follow closely. In the other regions, female secondary school enrollments are by far the most potent force for reducing child malnutrition. In all regions except the Near East and North Africa (NENA), access to safe water is the determinant that needs to change the most, relative to its range, to bring about a 1 percentage-point reduction in child malnutrition.

While the numbers in column (3) provide a sense of the relative strength of impact of the determinants, they say nothing about their current distance from desirable levels and hence their scope for bringing about reductions in child malnutrition over the medium-to-long run. The percent that each determinant is below its desirable level (in scale-neutral terms) is given in column (4). The desired levels of safe water access and female secondary school enrollments are assumed to be 100 percent. The desired level of the female-to-male life expectancy ratio is set at 1.01; that of per capita dietary energy supplies is set at 3,100. 21 Column (5) gives the estimated reduction in the prevalence of child malnutrition if each determinant were raised to its desirable level. This number provides a measure of potential contribution.

21 The desired level of the female-to-male life expectancy ratio is determined as the average of the top 20 percent of the data points in the panel data set, excluding the maximum value (1.15) of El Salvador in 1988, which is an extreme value compared to the other high ratios. There is no widely accepted “desirable” level of per capita DES from the standpoint of nutritional health. Countries with very high calorie levels also have high levels of obesity, an undesirable trait. For example, in 1995, Western Europe had an average DES of 3,360 (FAO 1998). This study estimates that at a DES of about 3,120 kilocalories, increases in dietary energy supplies no longer serve to reduce child malnutrition levels. Alexandratos (1995) says that 10 percent of a country’s population will be undernourished (or food insecure) at DES levels of 2,700 kilocalories. On the other hand, FAO (1996) claims that at a level of about 2,770 only 2.5 percent of the population will be undernourished. An intermediate level of 3,100 kilocalories was chosen here.

In all regions, increasing female secondary school education to its desirable level has the largest medium- to long-term potential to reduce child malnutrition. Food availability is second in Sub-Saharan Africa and LAC. Women’s relative status is second in South Asia, East Asia, and NENA.

To identify policy priorities for future reductions in child malnutrition for each region, the determinants are ranked in terms of the size of the change required to bring about a 1 percentage-point reduction in child malnutrition as a percentage of the determinants’ ranges (based on Table 14, column 3); and their potential for reducing it in the medium-to-long term (based on Table 14, column 5). Combining these two sets of ranks, the best estimates (in the absence of cost data) of future policy priorities for addressing the underlying determinants of child malnutrition in each developing region are summarized in Table 15.

In South Asia and Sub-Saharan Africa the top priorities are raising per capita food availability and women’s education. In both regions, improvements in per capita food availability have the strongest effect, but women’s education also has a strong effect, and it would make the biggest difference if increased to its desirable level. In East Asia, NENA, and LAC, women’s education is the top priority, both from the standpoint of strength of impact and scope for reducing child malnutrition. In East Asia, food availability and women’s relative status should also receive high priority. In NENA, women’s relative status is the second highest priority. In LAC, women’s relative status and health environment improvements tie for second. For South Asia, a secondary priority is improving women’s status relative to men’s, which, because it is so far below desirable levels, has great scope for reducing malnutrition.

Table 15 - Priorities by region for future child malnutrition reduction (underlying-determinant variables)


Region

Rank of determinants by most potent impact on malnutrition relative to its existing range
(1)

Rank of determinants by most potential for impact based on increases to desirable levels
(2)

Top Priorities
(3)

South Asia

1. Food availability

1. Women’s education

1. Food availability


2. Women’s education

2. Women’s relative status

1. Women’s education


3. Women’s relative status

3. Food availability

2. Women’s relative status


4. Health environment

4. Health environment


Sub-Saharan Africa

1. Food availability

1. Women’s education

1. Food availability


2. Women’s education

2. Food availability

1. Women’s education


3. Women’s relative status

3. Health environment



4. Health environment

4. Women’s relative status


East Asia

1. Women’s education

1. Women’s education

1. Women’s education


2. Food availability

2. Women’s relative status

2. Food availability


3. Women’s relative status

3. Health environment

2. Women’s relative status


4. Health environment

4. Food availability


Near East and North Africa

1. Women’s education

1. Women’s education

1. Women’s education


2. Women’s relative status

2. Women’s relative status

2. Women’s relative status


3. Health environment

3. Health environment



4. Food availability

4. Food availability


Latin America and the Caribbean

1. Women’s education

1. Women’s education

1. Women’s education


2. Women’s relative status

2. Health environment

2. Women’s relative status


3. Food availability

3. Food availability

2. Health environment


4. Health environment

4. Women’s relative status


Notes: The rankings in column (1) are based on the numbers reported in Table 14, column (3). The rankings in column (2) are based on the numbers reported in Table 14, column (5). The top priorities in column (3) are based on the highest ranked determinants in columns (1) and (2).

Box 2

The (South) Asian Enigma

In South Asia, 50 percent of the children under age five are malnourished; in Sub-Saharan Africa, 31 percent. Why is malnutrition so much higher in South Asia? The huge difference has been called an “enigma” because South Asia as a region is doing much better than Sub-Saharan Africa for most of the determinants of child malnutrition (see the table below) (Ramalingaswami, Jonsson, and Rohde 1996). There are three possible sources of these differences.

First, the determinants of child malnutrition may be different or have different strengths of impact in the regions. If one determinant is more important in South Asia than in Sub-Saharan Africa, and South Asia is not doing well in that area, then that determinant could be a clue to the enigma. This report finds no major differences in the importance of the underlying-determinant causal factors between the regions. For the basic determinants, some structural differences were evident, but it was not possible to find out which determinant, national income or democracy, was causing the difference.

The second possible source of the difference in child malnutrition rates may be that South Asia is doing worse than Sub-Saharan Africa in the factors studied. As the table shows, South Asia is doing better than Sub-Saharan Africa for all factors except women’s status relative to men’s. Therefore, it seems likely that women’s status is one reason for the higher prevalence of malnutrition in South Asia. The table also indicates that South Asia’s poverty rate is slightly higher than Sub-Saharan Africa’s, which may explain some of the difference.

Progress in some determinants of child malnutrition in South Asia and Sub-Saharan Africa, 1995


South Asia

Sub-Saharan Africa

Child malnutrition (percent)

49.3

31.1

Access to safe water (percent)

79.7

48.8

Female secondary school enrollment (percent)

34.2

19

Female-to-male life expectancy ratio

1.023

1.054

Per capita dietary energy supply (kilocalories)

2,356

2,136

Per capita national income (PPP US$)

1,136

778

Democracy

4.10

2.44

Poverty (percent) a

43.1

39.1

Sources: Smith and Haddad 2000, Tables 1, 25, and 26. Poverty data are from Ravallion and Chen 1996, Table 5.

Notes: With the exception of the poverty rates, these data are population-weighted means over all countries in the data set in each region. The poverty measure employs an international poverty line of $1 per person per day at 1985 purchasing power parity.

a Poverty figures are for 1993.

The final source of the difference in child malnutrition rates of the two regions lies in the “black box” of time-invariant, country-specific factors. Because the data set covers more than one point in time for each country, the effects on child malnutrition of these factors can be estimated, even though it is not possible to determine what they actually are. The factors are found to raise the prevalence of child malnutrition in South Asia well above Sub-Saharan Africa’s. To illustrate their importance in the regional differences, the figure below shows how much child malnutrition would remain, if all of the underlying-determinant variables were to reach their desirable levels. In South Asia, malnutrition would remain at 23.8 percent, but it would be only 0.5 percent in Sub-Saharan Africa. Deeply entrenched factors specific to South Asian countries, then, are also key to solving the Asian enigma. In the long run, if child malnutrition is to be overcome in the region, the black box must be opened to find out what these factors are and to implement policies to address them. Some possibilities are the monsoon climate (FAO 1996), recurrent flooding in some countries, overcrowding due to high population density, and cultural beliefs and traditions that hinder optimal breast feeding and timing of the introduction of complementary foods (Ramalingaswami, Jonsson, and Rohde 1996).


Predicted reductions in child malnutrition and remaining prevalence if underlying-determinant variables reach desirable levels

Note: At 0,5 percent, remaining malnutrition in Sub-Saharan Africa is too small to show on the figure.

Health environment improvements, in a relative sense appear to be a weak force for reducing child malnutrition. This determinant’s ranking is low partially because substantial progress has already been made in this area in many regions, compared with other determinants (Table 14, column 4). In an absolute sense it still makes a big difference, however. If universal access to a proper health environment (proxied by safe water) were achieved, the prevalence of child malnutrition would fall by 2.3 percentage points. The numbers of malnourished children would fall by 11.9 million.

Relative Importance of National Income and Democracy

Democracy is important in facilitating health environment improvements and increases in food availability. Per capita national income is important in maintaining and improving investments in health environment, female education, women’s relative status, and per capita food availability, both from the viewpoint of public investments and (through its association with household incomes) investments at the household level. Because strong regional differences in the effects of the basic determinants have been detected, this discussion is limited to their relative importance to the developing countries as a group. A comparison of their strengths and potential impacts is given in Table 14.

At this point in time, raising national incomes would have a stronger effect than enhancing democracy. It would take an increase of $202 in the average national per capita GDP of developing countries to reduce the prevalence of child malnutrition by 1 percentage point, which is 9.7 percent of its range (column 2). By contrast, it would take almost a 0.8 point rise in the democracy index to bring about the same reduction (11.5 percent of its range). The developing-country per capita GDP is currently far below any desirable number. Past a level of about $4,750 the factor loses its force in reducing child malnutrition. Even bringing the developing-country GDP up to this moderate level would have quite a large impact on child malnutrition: its prevalence is predicted to fall by 18.5 percentage points, the number of malnourished children by almost 100 million. The regions that have the longest way to go to reach the $4,750 mark (and thus the most to gain from doing so) are Sub-Saharan Africa, South Asia, and East Asia. Latin America and the Caribbean, as a region, has already surpassed the mark.

Relatively speaking, democracy is not a very strong force in reducing child malnutrition in the developing countries; nevertheless, improving it would make a big absolute difference to child malnutrition. If the democracy index were raised to its desired level (of 7), the prevalence of child malnutrition in the developing countries would fall by 5.5 percentage points. The numbers of children who are malnourished would be reduced by 29.4 million. The regions that have the longest way to go to reach a desirable level of democracy are East Asia, NENA, and Sub-Saharan Africa.

The reader should bear in mind that improvements in national income and democracy only lead to reductions in child malnutrition if they are directed to improvements in the underlying determinants. Given enhanced political will and education, it is possible that they can be even more effectively directed toward them in the future than they have in the past. This analysis gives governments an idea of where to best direct increased national incomes in the interest of children’s nutrition. It also suggests target areas where political will and commitment among democratic governments can be instilled.

A Note on Cost-Effectiveness

On a final note, a full assessment of priorities for the future should ideally take into account the costs of improving the alternative determinants. While the same reduction in child malnutrition could be brought about by a 13.1 percentage-point increase in access to safe water as a 4.6 percentage-point increase in female secondary school enrollment, it is unlikely that these increases would cost the same. Their costs are likely to differ by region and over time. Unfortunately, good quality comparative information on the cost-effectiveness of these different policies is lacking. One can still get a sense of how different the relative costs have to be before the conclusions reached on priorities are altered. For example, if it cost more than 2.8 times as much to increase female secondary school enrollment by 1 percent, compared with the costs of increasing access to safe water by 1 percent, then the latter will be the more cost-effective investment in reducing malnutrition. If the ratio is below 2.8, then female secondary education becomes more cost-effective. Better information on cost-effectiveness should be a key focus of future research on policies to improve child nutritional status.