4. New Evidence from Cross-Country Data, 1970-95

Descriptive Analysis

The regional levels and trends of underweight prevalence in the 63-country sample closely follow those for the developing countries as a whole in Table 1. South Asia had the highest prevalence throughout the period, roughly double that of the region in second place, Sub-Saharan Africa (Table 4). More than half of all South Asian children under five years old were underweight for their age. Roughly one-third were underweight in Sub-Saharan Africa and one-fifth in East Asia. The proportion of underweight children was smaller in Near East and North Africa (NENA) and Latin America and the Caribbean (LAC). The regions whose child malnutrition rates declined the most from the 1970s to the 1990s are South Asia and East Asia. Sub-Saharan Africa is the only region for which underweight rates have increased during the period.

Turning to the underlying-determinant explanatory variables, NENA and LAC had the highest access to safe water, over 70 percent, while Sub-Saharan Africa had the lowest, at 37.5 percent, which amply illustrates the high degree of inequality across the regions (Table 4, column 2). Access to safe water has improved substantially during the study period. It has more than doubled for the full sample, starting at 36 percent in the 1970s, increasing quickly to 62 percent in the 1980s, and rising to 69 percent by the 1990s (Table 5). Rates of improvement were greatest for East Asia and South Asia.

With respect to women’s education, at 16 percent, Sub-Saharan Africa had the lowest rate of enrollment of females in secondary schools (Table 4, column 3). The rate was also low in South Asia, where only 24 percent of women of eligible age were enrolled. For the entire sample, female secondary school enrollment rates improved steadily, rising from 22 percent in the 1970s to 45 percent in the 1990s. Nevertheless they are still quite low: less than half of the women in developing countries complete a secondary school education.

The measure of women’s status relative to men’s, the ratio of female-to-male life expectancy, was by far the lowest in South Asia, with men’s and women’s life expectancies being roughly equal (Table 4, column 4). Women’s life expectancy in the developed countries is on average six to seven years longer than men’s (Mohiuddin 1996). The ratio of women’s life expectancy to men’s in Norway, for example, is 1.08. Thus South Asia’s ratio of 1.01 is extremely low. Sub-Saharan Africa, East Asia, and NENA had ratios of 1.06, 1.05, and 1.04, respectively-rates well below those of developed countries. LAC had the highest ratio of the developing-country regions, which, at 1.09, is on a par with the developed countries. The ratio has improved or remained fairly steady in all regions except Sub-Saharan Africa. Over time, the ratio for the developing countries as a whole increased from 1.02 in the 1970s to 1.05 in the 1990s (Table 5).

Per capita dietary energy supplies (DES) were lowest in South Asia and Sub-Saharan Africa over the study period. The minimum daily dietary energy requirement for an active and healthy life is about 2,150 kilocalories (FAO 1996). These regions’ supplies (not intake) barely surpassed this requirement (Table 4, column 5). The minimum DES necessary (but not sufficient) for bringing food insecurity to a low 2.5 percent of a country’s population is 2,770 kilocalories (FAO 1996). The DES of East Asia and LAC neared this level during the study period; NENA’s surpassed it. From the 1970s to the 1990s DES increased in all regions except Sub-Saharan Africa.

Table 4 - Regional comparison of child malnutrition prevalences and explanatory variable means, 1970s to 1990s

Region/decade		Child malnutrition (1)	Access to safe water (2)	Female secondary school enrollment (3)	Female-to-male life expectancy ratio (4)	Per capita dietary energy supply (5)	Per capita GDP (6)	Democracy (7)
		(percent)	(percent)	(percent)		(kilocalories)	($PPP)	(1 = least democratic)
South Asia		61.0	60.5	23.8	1.010	2,187	863	4.59
	1970s (n = 4)	69.1	29.8	16.3	0.987	2,023	728	4.38
	1980s (n = 6)	61.8	51.9	14.2	1.020	2,042	719	3.25
	1990s (n = 6)	55.7	81.3	31.5	1.022	2,332	990	5.16
Sub-Saharan Africa		31.0	37.5	15.6	1.061	2,164	879	2.57
	1970s (n = 10)	27.2	24.7	8.5	1.069	2,207	1,358	1.77
	1980s (n = 26)	26.5	35.0	14.6	1.066	2,117	1,031	2.02
	1990s (n = 29)	33.7	40.4	17.0	1.060	2,184	740	2.96
East Asia		23.0	64.5	47.9	1.051	2,595	1,874	1.69
	1970s (n = 2)	45.0	19.7	25.8	1.050	2,007	1,402	3.0
	1980s (n = 13)	26.8	63.8	39.2	1.053	2,502	1,483	2.30
	1990s (n = 11)	19.4	67.8	54.4	1.049	2,686	2,132	1.25
Near East and North Africa		11.0	75.5	52.5	1.043	3,058	2,527	2.81
	1970s (n = 3)	16.5	72.5	34.0	1.042	2,710	1,547	3.32
	1980s (n = 4)	10.1	69.3	46.4	1.043	3,018	2,746	3.09
	1990s (n = 7)	10.8	79.4	59.7	1.043	3,157	2,637	2.55
Latin America and the Caribbean		12.0	71.8	44.8	1.094	2,647	4,740	4.73
	1970s (n = 12)	18.9	59.5	33.3	1.086	2.620	4,713	4.06
	1980s (n = 26)	11.4	79.0	47.2	1.096	2,675	4,871	5.14
	1990s (n = 20)	8.3	73.3	51.4	1.098	2,636	4,607	4.79

Notes: The means reported in this table are calculated based only on the country-year pairs included in the study data set. They are population-weighted. The numbers for the regions (in bold) are for the entire time period.

Table 5 - Variable means for the 1970s, 1980s, and 1990s, and correlations with the underweight rate

Variable	1970s	1980s	1990s	Change 1970s to 1990s	Percent change 1970s to 1990s	Correlation with under weight rate a
Child malnutrition (percent)	50.7	29.0	28.5	-22.2	-43.8	...
Access to safe water (percent)	36.3	61.6	69.0	32.7	+90.0	-0.50
Female secondary school enrollment (percent)	21.7	34.5	45.0	23.3	+107.0	-0.48
Female-to-male life expectancy ratio	1.024	1.055	1.047	0.023	+2.3	-0.43
Per capita dietary energy supply (kilocalories)	2,187	2,440	2,564	377	+17.2	-0.52
Per capita GDP ($)	1,772	1,871	1,904	132	+7.5	-0.59
Democracy (1 = least democratic)	3.96	2.86	2.66	-1.3	-32.8	-0.31
Number of observations	31	75	73	...	...	...
Number of countries	29	54	58	...	...	...

Note: The means reported in this table are calculated based only on the country-year pairs included in the study data set and therefore must be considered illustrative. They are population-weighted.

^a Pearson correlation coefficients. All are significant at the 1 percent level.

South Asia and Sub-Saharan Africa had the lowest per capita national incomes and LAC the highest by far (Table 4, column 6). The only region that experienced negative growth was Sub-Saharan Africa. For the developing countries as a whole, per capita GDP increased by about 7 percent between the 1970s and the 1990s.

The region that has been least democratic is East Asia (Table 4, column 7). Interestingly, South Asia and LAC, while at opposite extremes in terms of underweight, were almost equally democratic over the 25-year period. These regions had the highest democracy index scores. Democracy has improved for South Asia, Sub-Saharan Africa, and LAC; it has deteriorated for East Asia and NENA. It is the only explanatory variable that has declined for the developing-country sample as a whole, with scores falling from about 4.0 in the 1970s to 2.7 in the 1990s (Table 5).

In the last column of Table 5, correlations between underweight rates and each explanatory variable are given. The correlation coefficients for all variables are negative and statistically significant, indicating fairly strong negative associations between child malnutrition and the hypothesized determinants. The variables with the strongest correlations are per capita national income, per capita DES, and access to safe water. The weakest correlation is for democracy. In the next section, multivariate analysis is used to single out the independent effects of changes in each variable, while controlling for the others.

Multivariate Analysis

Table 6 reports fixed-effects regression results for the underlying determinant and basic determinant models. The coefficients of all of the underlying determinants are statistically significant and negative (column 1). Increased access to safe water (as a proxy for the health environment), increased education and improved status for women (as proxies for maternal and child care and for food security), and increased quantities of food available at a national level (for food security) all work to reduce levels of child malnutrition in developing countries. While per capita food availability has a negative relationship with child malnutrition, it has a declining marginal effect: at low levels of per capita DES, the relationship is strongest; as levels increase, its effect weakens. This relationship is captured by estimating a regression coefficient for three different ranges of DES. The first segment (DES less than 2,300 kilocalories) has a relatively large coefficient. The coefficient of the last segment (DES greater than 3,120 kilocalories) is not statistically significant, implying that after reaching a level of about 3,100 kilocalories, further increases in per capita DES no longer contribute to reductions in child malnutrition.

Table 6 - Country fixed-effects estimation results

Variable		Underlying determinants	Basic determinants
Access to safe water		-.076 (1.95)*	...
Female secondary school enrollment		-.220 (3.41)***	...
Female-to-male life expectancy ratio		-71.8 (1.74)*	...
Per capita dietary energy supply (DES)
	DES £ 2,300	-.0170 (3.41)***	...
	2,300 < DES £ 3,120	-.0024 (2.16)**	...
	DES >3,120	.0405 (1.35)	...
Per capita GDP (GDP)
	GDP £ 800	...	-.0444 (3.15)***
	800 < GDP £ 4,725	...	-.0067 (2.63)***
	GDP > 4,725	...	.0006 (3.37)***
Democracy		...	-1.27 (2.51)**
R2		.947	-.930

Notes: The dependent variable is prevalence of child malnutrition measured as the percent of underweight children under age five. The coefficients on the fixed-effects terms are not shown. The number of observations for all regressions is 179 (63 countries). Absolute values of t-statistics are given in parentheses.

* Significant at the 10 percent level,

** Significant at the 5 percent level.

*** Significant at the 1 percent level.

Turning to the basic determinants model, the regression coefficients on both per capita GDP and the democracy index are statistically significant and negative (column 2). Both increases in overall income at a national level (regardless of how distributed) and the extent of democracy serve to reduce child malnutrition in developing countries. Per capita GDP exhibits a declining marginal effect on child malnutrition: once $4,700 per capita is passed (which few countries in the sample surpass), further increases in per capita GDP no longer contribute to reductions in child malnutrition, as indicated by a near zero (yet statistically significant) coefficient.

Table 7 - Underlying-determinant variable regressions with basic-determinant variables as independent variables

Variable	Access to safe water (1)	Female secondary school enrollment (2)	Female-to-male life expectancy ratio (3)	Per capita dietary energy supply (4)
Per capita GDP	.0174 (2.85)***	.0148 (3.71)***	1.0E-05 (1.90)*	.4105 (6.26)***
GDP2	-1.31E-06 (2.31)**	-9.32E-07 (2.53)**	-8.2E-10 (1.67)*	-2.79E-05 (4.59)***
Democracy	3.49 (2.87)***	.981 (1.23)	-.002 (1.57)	26.28 (2.0)**
R2	.835	.922	.901	.902
Adjusted R2	.740	.877	.845	.846

Notes: The number of observations for all regressions is 179 (63 countries). Absolute values of t-statistics are given in parentheses. The regressions are estimated using a country fixed-effects specification.

* Significant at the 10 percent level.

** Significant at the 5 percent level.

*** Significant at the 1 percent level.

According to the conceptual framework (Figure 1), the basic determinants affect child malnutrition through their influence on the underlying determinants. Fixed-effects regression estimates of the effects of per capita national income and democracy on each underlying determinant are given in Table 7. The per capita GDP coefficients are positive and significant for all of the underlying determinants. The results indicate that the amount of income available per person in a country is an important resource base for investment - both public and private - in health environments, women’s education, women’s relative status, and national food availabilities. ¹¹ The coefficient of the democracy index in the safe water access and per capita dietary energy supply equations is significant and positive. Therefore, it seems probable that democratic governments are more likely to direct their budgets to improvements in health environments and food availabilities than nondemocratic ones. They are not more likely to direct public resources toward women’s education or to women vis-is men.

¹¹ significant and negative GDP-squared term for all of the underlying determinants implies that the impact of incremental increases in per capita national income tends to decline as incomes rise.

How substantial, in a practical sense, are the estimated effects of the underlying and basic determinants on child malnutrition, and how do they compare across determinants? It is difficult to get a sense of the relative strengths of the variables’ effects just by looking at their regression coefficients. This is because each variable is measured in different units and has a different range. Table 8 translates the results into more meaningful terms. Column (2) gives the increase in each variable that would be required to bring about the same reduction in the prevalence of child malnutrition: one percentage point. Each determinant’s range, based on the minimum and maximum values observed among developing countries during 1970-95 is given in column (3), while column (4) gives the number in column (2) as a percent of the determinants’ ranges, which is a scale-neutral measure of strength of impact. ¹²

¹² A more common measure is elasticity, which gives the percentage change in an outcome variable associated with a 1 percent increase in an explanatory variable. This measure is not used here because it does not account for the fact that the variables being compared have different ranges. It thus gives an inaccurate ranking of their strengths of impact relative to one another. For example, the elasticity of child malnutrition with respect to the female-to-male life expectancy ratio is -3.1, while the elasticity with respect to female secondary school enrollment is -0.17. However, a 1 percent increase in the former represents a very large increase in it (7 percent of the total range) compared to a 1 percent increase in the latter (0.34 percent of the range). Their elasticities, while informative, are thus not comparable.

Table 8 - Comparison of the effects of variables on child malnutrition

Variable			Sample(or segment) mean (1)	Increase in variable needed to reduce prevalence of child malnutrition by 1 percentage point a (2)	Developing-country range (3)	Number in (2) as a percent of developing-country range (4)
Underlying-determinant variables
	Access to safe water (percent)		56.2	13.1	1-100	13.2
	Female secondary school enrollment(percent)		33.8	4.6	0.5-100	4.6
	Female-to-male life expectancy ratio		1.0624	0.0139	0.97-100	9.3
	Per capita dietary energy supply (DES) (kilocalories)		2,360	101	1,522-3,605	4.9
		DES £ 2,300	2,106	59	...	2.8
		2,300 < DES £ 3,120	2,613	425	...	20.4
		DES > 3,120	3,230	...	...	...
Basic-determinant variables
	Per capita GDP ($)		2,306	74.1	300-8,612	0.89
		GDP £ 800	645	23	...	0.3
		800 < GDP £ 4,725	2,102	150	...	1.8
		GDP > 4,725	5,867	...	...	...
	Democracy		3.5	0.79	1-7	13.1

Note: Leaders indicate not applicable.

^a Calculated as 1 divided by the regression coefficients of Table 6.

The estimates indicate that a 13.1 percentage point increase in population with access to safe water would be required to bring about a 1 percentage point reduction in the child malnutrition prevalence. This represents 13.2 percent of the variable’s range. By contrast, the required increase in the female secondary school enrollment rate is only 4.6 percentage points, representing only 4.6 percent of its range. Thus the required increase in safe water access to bring about the same reduction in child malnutrition is much higher than the required increase in female enrollments, implying that increases in secondary education for women are likely to have a stronger (more negative) impact on child malnutrition than are increases in access to safe water.

The required increase in per capita DES for the full sample (101 kilocalories) is 4.9 percent of its range; that of the female-to-male life expectancy ratio (0.0134) is 9.3 percent of its range. Therefore a rough ranking of the underlying determinants in terms of their potency in reducing child malnutrition is: women’s education (greatest potency), followed closely by per capita food availability, followed, third, by women’s relative status, and fourth by health environment improvements. Note that for the low DES range (£ 2,300 kilocalories), per capita food availability would be ranked first and women’s education second. For the medium and high DES ranges (>3,120), however, women’s education would be ranked first and per capita food availability last. The policy implications of these rankings will be drawn out more fully in the conclusions (Chapter 8).

For the basic determinants (lower panel of Table 8), per capita national income appears to be a more potent force for reducing child malnutrition than democracy. The required increase in GDP per capita to reduce the prevalence of child malnutrition by 1 percentage point is $74. This is less than 1 percent of the variable’s range, a small proportion. In contrast, a very large increase in democracy would be required to bring about the same change: an increase in the index of 0.8 points (13 percent of its range). Per capita national income has a stronger impact than democracy, even for the medium GDP group (between $800 and $4,725). For the high GDP group (>$4,725), however, democracy prevails as the most potent basic determinant because national income has only a minor impact on child malnutrition in this range.

How accurate are the regression estimates reported in Table 6? Concerns about incorrect or “biased” parameter estimates as the result of either omission of relevant explanatory variables or endogeneity problems appear to be unfounded. Both the underlying- and basic-determinant models pass a test for the absence of bias from omitted variables, suggesting that the major factors determining child malnutrition at these levels of causality have successfully been captured. Instrumental variables (IV) tests for endogeneity of all variables were undertaken, with the exception of the female-to-male life expectancy ratio and the democracy index, which were not tested because of data constraints (Smith and Haddad 2000). The test results indicate that health environment quality, women’s education, per capita food availability, and per capita national income are not endogenous in the empirical models of child malnutrition specified. Given these test results, it is assumed that the estimates are as accurate as possible given current data limitations. In addition, that the estimations are based on a sound conceptual framework (Figure 1) and are undertaken with respect to changes over time in the variables provides further assurance that a causal, rather than merely associative, relationship between child malnutrition and the explanatory variable has been identified.

Past studies suggest that there may be differences across the developing regions in the determinants of child malnutrition or in the magnitude of their effects, especially for South Asia. A test for regional differences in the estimates for underlying-determinant coefficients identifies no strong differences. Thus it is assumed that the estimates in Table 6, column (1) apply to all of the regions. Whereas, from a structural standpoint, the relationship between child malnutrition and DES does not differ substantially across the regions, the regions do differ greatly in the levels of their per capita DESs. Because the strength of this determinant depends on its level, the regions thus differ greatly in the strength of impact of DES on child malnutrition. Corresponding to their low per capita DESs over the study period, the effects for Sub-Saharan Africa and South Asia are the highest in magnitude. The other regions have substantially higher DES per capita, and thus their coefficient estimates are much lower in magnitude.

For the basic determinants, test results suggest that there are structural differences across the regions in the effects of national income or democracy or both. South Asia, in particular, differs fundamentally from the others. As for per capita food availability, the effect of per capita national incomes on child malnutrition for any region depends on its level. In South Asia and Sub-Saharan Africa, which had the lowest per capita GDPs during the study period, the effect of national income is relatively strong. It is much weaker for East Asia, NENA, and LAC.

The final clue as to whether substantial regional differences exist in the causes of child malnutrition lies in the magnitudes of the country fixed-effects terms included in the regression equations. These terms represent the effects of factors that have not changed much (over approximately 13-year periods, the average time span covered for a country). A clear result from the analysis is that the influence on child malnutrition of these unobserved factors is much stronger for South Asia than for the other regions. The mean of the fixed-effects coefficients in the underlying-determinants model is 9.6. This means that, independent of the levels of the explanatory variables included in the regression equation, the prevalence of child malnutrition in the developing world would be about 10 percent. The mean of the fixed-effects coefficients for South Asian countries is far above that of the sample and the other regions, at 33.3.

A Cautionary Note

Because this study employs cross-cutting empirical methods, its results apply only at the very broad level of the developing countries as a whole and, more tentatively, to the developing-country regions. Their applicability to specific populations at more disaggregated levels is unknown. Careful analysis and diagnosis are needed to understand the causes of child malnutrition for each subpopulation of the developing world, whether it be a region, a country, an area within a country, a community, a household, or an individual child.