RM Project Final 30.3

c c
RESEARCH METHODOLOGY
Research Report on:

m m
mm m m m
Submitted To:
Dr. Vijay Wagh
Div-B
Prepared By:
Nishi Avasthi 239
Ravi Agarwal 244
Swati Agrawal 253
Vaishali Radhakrishnan 255
N. L. Dalmia Institute of Management Studies and Research c

c
c
c c
PART A
PREFATORY ITEMS

c
c
c c
Letter of Transmittal

Dear Sir,
As per your directions in Letter of Authorization, we have completed our research

on ´To study the increasing rate of malnutrition in India D c The report is based on
interviews and responses of respondents from Mumbai. The project was carried

from c
c cc c c .The complete methodology and conclusions
derived on the basis of the responses is described in the report. We believe you will
find the results mentioned in the report to be interesting and certainly useful.
Sincerely,
Nishi Avasthi
Ravi Agarwal
Swati Agrawal
Vaishali Radhakrishnan
Div B ² Group

c
c
c c
RESEARCH METHODOLOGY
Research Report on:

m m
mm m m m
Submitted To:
Dr. Vijay Wagh
Div-B
Prepared By:
Nishi Avasthi 239
Ravi Agarwal 244
Swati Agrawal 253
Vaishali Radhakrishnan 255

c
c
c c
Letter of Authorization
c c
c c c c c c c c c c
Subject: Letter of authorization to conduct research
Dear Students,
I have strong conviction that you are going to achieve a decision making position in
your career within a short span of time. Let it be any function of the organization,
unless one has a realistic picture and approach of the thriving factors of any decision
to be made, it becomes extremely difficult to arrive at a decision which is equally
risky. Thus the study of research methodology will provide you with the knowledge
and skills and help you learn different techniques of conducting research which form
a very important part of your job.
Thus to have a deeper insight in research methodology, I would like you to study
´To study the increasing rate of malnutrition in Indiaµ and hope it will help you to
study the process research and hope it will help you to know procedures of
collecting data, preparing questionnaires, preparing reports, etc.
Yours sincerely
!"
Core Faculty, Marketing
NLDIMSR
Mumbai.

c
c
c c
Acknowledgement
We take this opportunity to thank all those individuals who helped in preparing this
report. We want to express our appreciation to all the travelers who commute
between town and suburb whose insightful comments and suggestions were
invaluable to us during the research. We would also like to express our sincere
gratitude to Dr. Vijay Wagh for giving us an opportunity to learn and explore the
reasons for the increasing rate of malnutrition in India and also guiding us
throughout our research.
Also we would like to thank Prof. P. L. Arya, Director, N. L. Dalmia Institute of

Management Studies and Research.

c
c
c c
#$%$&&
The main purpose of conducting this research is to study and analyze the causes for
the increasing rate of malnutrition in India. This study will help us understand the
lifestyle of the people in rural & urban areas & also help to understand the degree of
awareness among people about malnutrition. In order to achieve our desired
objectives, we have collected data from the government records and social human
rights association, etc.
This study gives a bird·s eye view of the malnutrition statistics & trends prevalent
among the Indian population (both rural & urban). To accomplish the task, a survey
will be carried in and around Mumbai city with a sample of 100 respondents that
will be on a random. A structured questionnaire is prepared and will be
administered and the data so collected will be analyzed both by percentages and
statistical methods. Extensive use of Microsoft Excel software will be incorporated
for getting the analytical report.
The interviews will be conducted and data will be collected accordingly. Standard
editing and coding procedures would be utilized to ensure maximum accuracy and
unambiguity. This includes careful interpretation and good judgment of the data.
Hypothesis testing will also be undertaken. Simple tabulation will be utilized to
analyse the data.
A written report will be prepared and the presentation of the findings will be made.
A schedule of the research program as prepared will be followed so as to complete
the research on time and in budgeted limits.

c
c
c c
'()*)++
Sr. No. Topic Page No.
,-,
.m
*+&( 3
(," 4
1
*$)/)+ 5
01+)2(3"&+ 6
4#$%$&& 7
,5-m m
10
,)'(&&+
11
2 '%
12
6 7)
13
$%2
,-6 .
"+ 112
&7(+",(+ 113
3
)(()+ 114
0+( 115
4&)+ 121
,- m
4 0
+3+" 122
0)+($)+ 123
5 6, , 124
6 ,-,, m 139

c
c
c c
Part B
INTRODUCTION

c
c
c c
, 5

To study malnutrition


c
c
c c
6 58m

i c To study the effect of lower income levels on malnutrition rate in India.
^ c To study the effect of inflation on malnutrition rate in India
u c To study the effect of poor eating habits on malnutrition rate in India.
º c To study the effect of inefficient Public Distribution System on malnutrition
rate in India
Ñ c To study the effect of inadequate storage facilities on malnutrition rate in
India

c
c
c c
6 7)
1. c H0 There is no significant effect of lower income levels on

malnutrition rate
H1 There is significant effect of lower income levels on
malnutrition rate
2. c H0 There is no significant effect of inflation on malnutrition rate.
H1 There is significant effect of inflation on malnutrition rate.
3. c H0 There is no significant effect of poor eating habits on

malnutrition rate.
H1 There is significant effect of poor eating habits on malnutrition
rate
4. c H0 There is no significant effect of inefficient Public Distribution

System on malnutrition rate
H1 There is significant effect of inefficient Public Distribution
5. c H0 There is no significant effect of inadequate storage facilities on

malnutrition rate
H1 There is no significant effect of inadequate storage facilities on
malnutrition rate

c
c
c c
m m!
!&(+$)+9
The term malnutrition generally refers both to under nutrition and over nutrition, but inthis
guide we use the term to refer solely to a deficiency of nutrition. Many factors cancause
malnutrition, most of which relate to poor diet or severe and repeated infections,particularly
in underprivileged populations. Inadequate diet and disease, in turn, areclosely linked to the
general standard of living, the environmental conditions, and whether a population is able
to meet its basic needs such as food, housing and health care. Malnutrition is thus a health
outcome as well as a risk factor for disease and exacerbated malnutrition (Fig. 1.1), and it can
increase the risk both of morbidity and mortality1 . Although it is rarely the direct cause of
death (except in extreme situations, such as famine), child malnutrition was associated with
54% of child deaths (10.8 million children) in developing countries in 2001 (Fig. 1.2; see also
WHO, 2004). Malnutrition that is the direct cause of death is referred to as ´protein-energy
malnutritionµ in this guide 2. Nutritional status is clearly compromised by diseases with an
environmental component, such as those carried by insect or protozoan vectors, or those
caused by an environment deficient in micronutrients. But the effects of adverse
environmental conditions on nutritional status are even more pervasive. Environmental
contamination (e.g. destruction of ecosystems, loss of biodiversity, climate change, and the
effects of globalization) has contributed to an increasing number of health hazards (Johns &
Eyzaguirre, 2000), and all affect nutritional status.
Overpopulation, too, is a breakdown of the ecological balance in which the population may
exceed the carrying capacity of the environment. This then undermines food production,
which leads to inadequate food intake and/or the consumption of non-nutritious food,
and thus to malnutrition. On the other hand, malnutrition itself can have far-reaching
impacts on the environment, and can induce a cycle leading to additional health problems
and deprivation.
1. Abler, D.G., G.S. Tolley and G.K. Kriplani, 1994: j

Westview Press: Boulder CO.
2. ACC/SCN (Administrative Committee on Coordination/Sub-Committee on Nutrition),
2003: 5

. Draft. United Nations.

c
c
c c
For example, malnutrition can create and perpetuate poverty, which triggers a cycle that
hampers economic and social development, and contributes to unsustainable resource use
and environmental degradation (WEHAB, 2002)3. Breaking the cycle of continuing poverty
and environmental deterioration is a prerequisite for sustainable development and survival.
(+$)++2)&++3(3+
The nutritional status of women and child ren is particularly important, because it isthrough
women and their off-spring that the pernicious effects of malnutrition arepropagated to
future generations. A malnourished mother is likely to give birth to a low birthweight
(LBW) baby susceptible to d isease and premature death, which only further undermines the
economic development of the family and society, and continues the cycle of poverty and
malnutrition 4. Although child malnutrition declined globally during the 1990s, with the
prevalence of underweight children falling from 27% to 22% (de Onis et al., 2004a), national
levels of malnutrition still vary considerably (0% in Australia; 49% in Afghanistan)(WHO,
2003).
The largest decline in the level of child malnutrition was in eastern Asia where und erweight
levels decreased by one half between 1990 and 20005 . Underweight rates also declined in
south-eastern Asia (from 35% to 27%), and in Latin America and the Caribbean the rate of
underweight children decreased by one third (from 9% to 6%) over the last 10 years. In
contrast, south-central Asia still has high levels of child malnutrition, even though the rate of
underweight children declined from 50% to 41% during the 1990s. In Africa, the number of
underweight children actually increased between 1990 and 2000 (from 26 million to 32
million), and 25% of all children under five years old are underweight, which signals that

2000a: º

. New York: United Nations, in collaboration
with the International FoodPolicy Research Institute, Washington D.C.
2000b,

º !" ###
$ %
& &
& .
5. Adair, L., 1999: Filipino children exhibit catch -up growth from age 2 to 12 years, "

%

,1291140-1148.

c
c
c c
little changed from a decade earlier6. The projection for 2005 is that the prevalence of child
malnutrition will continue to decline in all regions but Africa, which is dominated by the
trend in sub-Saharan Africa (de Onis et al., 2004b)7.
6.Adato, M., R. Meinzen-Dick, L. Haddad, P. Hazell, 2003, Impacts of Agricultural Research

on PovertyReduction: Findings of an Integrated Economic and Social Analysis, IFPRI:
Washington, DC.
7. Aduayom, D., and L. Smith, 2003: Estimating undernourishment with household
expenditure surveys: Acomparison of methods using data from three sub-Saharan African

c
c
c c

countries. In G
' % %
( ÿ
FAOÿ Rome.c
c
c
c
c

N. L. D li In i of Managemen Studies and Research c
c
c
c c
Man factors can contribute to high rates of child malnutritionÿ ranging from those
asfundamental as political instability and slow economic growthÿ to highly specific onessuch
as the frequency of infectious diseases and the lack of education. These factors canary
across countries. A cross-country analysis found that the determinants of stunting in
preschool children varied considerably between nationsÿ and among provinces within
nations (Frongillo et al.ÿ 1997). Important determinants of child malnutritionÿsuch as the

prevalence of intrauterine growth retardation (I R)ÿ also differ considerably across
geographicalregions(de Onisÿ Blössner & Villarÿ 1998). Whether or not children are
undernourished therefore seems to be as much a consequence of national and provincial
factorsÿ as of individual and household circumstances.

c
c
c c

(+$)++3(3")2
Malnutrition commonly affects all groups in a community, but infants and young children
are the most vulnerable because of their high nutritional requirements for growth and
development. Another group of concern is pregnant women, given that a malnourished
mother is at high risk of giving birth to a LBW baby who will be prone to growth failure
during infancy and early childhood, and be at increased risk of morbidity and early death8.
Malnourished girls, in particular, risk becoming yet another malnourished mother, thus
contributing to the intergenerational cycle of malnutrition.
In developing countries, poor perinatal conditions are responsible for approximately 23% of
all deaths among children younger than five years old (Fig. 1.2). These deaths are
concentrated in the neonatal period (i.e. the first 28 days after birth), and most
areattributable to LBW (Kramer, 1987)9. LBW can be a consequence of IUGR, pretermbirth,
or both, but in developing countries most LBW births are due to IUGR (defined asbelow the
tenth percentile of the Williams sex-specific weight-for-gestational agereference data).
Although the etiology of IUGR is complex, a major determinant ofIUGR in developing

countries is maternal under nutrition. Evidence has shown that there is a greater incidence
of IUGR births among women who are underweight or stuntedprior to conception, or who
fail to gain sufficient weight during pregnancy (Kramer, 1987; King & Weininger, 1989;
WHO, 1995a; Bakketeig et al., 1998), compared to women with normal weight and weight
gain.
8. Ahn, N. and A. Shariff, 1995: Determinants of Child Height in Uganda: A Consideration of

the SelectionBias Caused by Child Mortality, a

16(1), 49-59.
9.Alaii, N., et al.., 2003: Perceptions of Bed Nets and Malaria Prevention Before and After
RandomizedControlled Trial of Permethrin-Treated Bed Nets in Western Kenya,

"

%j
, Î (Suppl), 142-48.

c
c
c c
Growth assessment is the single measurement that best defines the health and nutritional
status of a child, because disturbances in health and nutrition, regardless of their etiology,
invariably affect child growth. There is ample evidence that the growth (height and weight)
of well-fed, healthy children from different ethnic backgrounds and different continents is
remarkably similar, at least up to six years of age (Habicht et al., 1974)10. Based on this
finding, WHO has been recommending that a single international reference population be
used worldwide, with common indicators and cut-offs, and that standard methods be used
to analyze child growth data (Waterlow et al., 1977; WHO, 1995a).11
Moreover, growth assessment is universally applicable: it does not pose any
culturalproblems; measuring equipment is easy to transport; the tools are simple and robust,
can be set up in any environment; users require little training; and the procedure is
inexpensive and non-invasive (WHO, 1995a).

0! &$3'$3+)*&(+$)+9
National estimates of the burden of malnutrition, including estimates for childmalnutrition,
provide vital information on preventable ill-health, and indicate the health gains possible
from interventions to prevent the risk factor (malnutrition, in this guide). 12 The results also
allow policy-makers to direct resources to the most vulnerable segments of the population,
and thus make better use of resources. Methods for estimating the burden of malnutrition
associated with poverty are outlined in the tenth volume of the EBD series (Blakely, Hales &
Woodward, 2004).13
10. Alauddin, M. and C. Tisdell, 1986: Market analysis, technical change and income
distribution in semi subsistence agriculture: The case of Bangladesh, )

: 1-18.
11. Alderman, H. and J. R. Behrman, 2003: Estimated Economic Benefits of Reducing LBW in
Low-Income Countries, University of Pennsylvania: Philadelphia, PA.
12. Alderman, H., J. Behrman, V. Lavy and R. Menon, 2001, Child health and School
Enrollment: A Longitudinal An alysis, "

%
Î(1): 185-205.
13. Alderman, H., J. Behrman, D. Ross and R. Sabot, 1996: The Returns to Endogenous
Human Capital in Pakistan·s Rural Wage Labor Market, %

% )

58(1), 29-55.

c
c
c c
To illustrate how to calculate the national burden of malnutrition, we give a step-by-step

numerical example for child malnutrition in Nepal, a developing country in the WHO SEAR
D sub region. For a complete listing of countries by WHO sub regions see Table 1 in the
annex. The disease burden is estimated in terms of the mortality and morbidity associated
with the principal causes of child death, i.e. diarrhea, acute respiratory infections, measles,
malaria, and perinatal risk factors; and with protein-energy malnutrition as a direct cause of
death, and mortality associated with other infectious diseases.
c )7)+()+7
Studies have demonstrated that the more malnourished children are, the sicker they areand
the higher their risk of early death (Pelletier, 1991; Toole & Malkki, 1992; Man etal., 1998)
(Fig. 2.1).
Severe malnutrition leads not only to increased morbidity(incidence and severity) and
mortality, but can also lead to impaired psychological andintellectual development. Growth
retardation in early childhood, for example, has beenlinked to the delayed acquisition of
motor skills (Heywood, Marshall & Heywood, 1991;Pollitt et al., 1994) (Fig. 2.2) and to
delayed mental development (Pollitt et al., 1993;Mendez & Adair, 1999).14 These outcomes
can have severe consequences in adult life,such as significant functional impairment
(Martorell et al., 1992; WHO, 1995a), that canaffect a person·s economic productivity. A
small adult may have a lower physical workcapacity than a larger adult, thus reducing
economic potential (Spurr, Barac-Nieto &Maksud, 1977); and small women in particular
may have obstetric complications (WHO, 1995a).15 Not surprisingly, malnutrition is closely
associated with socioeconomic status variables such as income and education.
14. Alderman, H., J. Hoddinott and B. Kinsey, 2003: Long Term Consequences of Early
Childhood
Malnutrition, World Bank: Washington, DC, Processed.44
15. Allen L., 2000: Anemia and Iron Deficiency: Effects on Pregnancy Outcome,
"

%$
, 71(5): 1280S-1284S.

c
c
c c
In all of these outcomes there is a doseresponse relationship with malnutrition. Specific

doseresponse relationships between impaired growth status and both poor school
performanceÿ as well as reduced intellectual achievementÿ are shown in Fig. 2.3 (see also
Martorell et al.ÿ 1992; PAHOÿ 1998). The evidence base for malnutrition as a risk factorÿ and
its dosedependent relationship with adverse health outcomesÿ are discussed in the
following sections for children (Section 2.1) and mothers (Section 2.2).

c
c
c c
c
cc&
c
ven though it has long been recognied that malnutrition is associated with
mortalityamong children (Trowellÿ 1948; Gomez et al.ÿ 1956)ÿ a formal assessment of the
impactof malnutrition as a risk factor was only recently carried out. In the early 1990sÿ
resultsof the first epidemiological study on malnutrition showed that malnutrition
c
c
c c
potentiatedthe effects of infectious diseases on child mortality at population level

(Pelletier,Frongillo & Habicht, 1993), a result that up until then had only been observed
clinically.The methodology was based on the results of eight community-based prospective
studies that looked at the relationship between anthropometry and child mortality in
developing countries (Pelletier et al., 1994)16 . The literature review used to select the eight
studies was published separately (Pelletier, 1994). The results of the eight studies were
consistent in showing that the risk of mortality was inversely related to weight-for-age, and
that there was an elevated risk even at mild-to-moderate levels of malnutrition. In fact, most
malnutrition-related deaths were associated with mild-to-moderate, rather than severe,
malnutrition, because the mild-to-moderately malnourished population was much bigger
than the severely malnourished population. The study also confirmed that malnutrition has
a multiplicative effect on mortality. Taking into account all underlying causes of death, the
results suggested that malnutrition was an associated cause in about one half of all child
deaths in developing countries.17
From a national policy perspective, however, the epidemiological study had a limitation: the
global estimate of malnutrition-associated mortality could not be applied to countries with
distinct disease profiles. To fill this gap, a joint WHO/Johns Hopkins University working
group was set up to estimate the contribution of malnutrition to cause-specific mortality in
children. The first step was a literature review to collect data for estimating the relationship
between malnutrition and mortality from diarrhea, acute respiratory infections, malaria and
measles (Rice et al., 2000). Cause-specific mortality was estimated by applying the method of
Pelletier et al. (1994) to the data of 10 cohort studies that contained weight-for-age categories
and cause-of-death information.
16. Allen, L., and S. Gillespie, 2001: What Works? A Review of Efficacy and Effectiveness of
Nutrition Interventions. Asian Development Bank, Nutrition and Development Series # 5,
Manila.
17. Alnwick, D., 1998: Weekly Iodine Supplements Work, "

% $

67(6): 1103-5.

c
c
c c
The weight for- age categories were based on the number of standard deviations (SDs) from
the median value of the National Centre for Health Statistics (NCHS)/WHO international
reference population (< -3 SD; -3 SD to < -2 SD; -2 SD to < -1 SD; and > -1 SD).18 All the
included studies contributed information on weight-for-age and risk of diarrhea, malaria,
measles, acute respiratory infections and all-cause mortality (comprising other remaining
infectious diseases besides HIV). These other infectious diseases include: tuberculosis,
sexually transmitted disease excluding HIV, pertussis, poliomyelitis, diphtheria, tetanus,
meningitis, hepatitis B and C, tropical-cluster diseases, leprosy, dengue, Japanese
encephalitis, trachoma, intestinal nematode infections, upper respiratory infections and
otitis media.19 By including these other infectious diseases, the burden estimates take into
account, for example, malnutrition associated effects on immune system and consequent
worsened prognosis of disease development.
The relationship between weight-for-age and risk of death was estimated by calculating the
logarithms of the mortality rates by cause and by anthropometric status for each country,
and using weighted random effects models. Using these models, the working group derived
the relative risks of dying for each cause and all causes. The detailed relative risks are
provided inTable 3.1, Section 3.4. The attributable fractions of mortality associated with a
weight-for-agelower than -1 SD from the median value were 44.8% for measles, 57.3% for
malaria, 52.3% for acute respiratory infections, 60.7% for diarrhea, and 53.1% for other
infectious diseases (all-cause mortality).
To estimate cause-specific morbidity, a statistical meta-analysis of published data
wasconducted to select longitudinal studies that compared incidence data according to
pastanthropometric status.20
18. Alston, J., C. Chan-Kang, M. Marra, P. Pardey and T.J. Wyatt, 2000: A meta-analysis of
rates of return to agricultural R&D, IFPRI Research Report No. 113, International Food
Policy Research Institute, Washington DC.
19. Altonji, J. and T. Dunn, 1996: The Effects of Family Characteristics on the Returns to
Education,
%)

78(4): 692-704.
20. Altonji, J. and C. Pierret, 2001: Employer Learning and Statistical Discrimination,
* "

%)

116(1): 313-350.

c
c
c c
According to the results, underweight status among preschool-age children was

significantly associated with subsequent risk of episodes of diarrhea and acute respiratory
infection, but the association with clinical malaria was not statistically significant. There was
no evidence that underweight status influenced susceptibility t o measles infection.21the
overall attributable fractions of morbidity associated with weight for-age below -2 SD were
16.5%, 5.3% and 8.2% for acute respiratory infections, diarrhea and malaria, respectively (see
Table 3.2, Section 3.5 for relative risks).
There is evidence that the burden of disease associated with malnutrition extends beyond
this approach, which looks only at certain disease groups. But until there are community
based cohort studies from which relative risks can be derived for other disea ses, an
assessment of the national burden of malnutrition will be restricted to these that were
studied.
+(&(+$)+
In 1995, WHO published a meta-analysis based on data sets from 25 studies that
relatedmaternal anthropometry to pregnancy outcomes (WHO, 1995b). The meta-
analysisreported that a pre-pregnancy bodymass index (BMI) below 20 kg/m2 was
associatedwith a significantly greater risk for IUGR, relative to a BMI above 24 kg/m2, with
anoverall odds ratioof 1.8 (95% confidence interval (CI): 1.7²2.0). In developingcountries, it
has been estimated that poor nutritional status in pregnancy accounts for14% of fetuses with
IUGR, and maternal stunting may account for a further 18.5%(ACC/SCN, 2000). Estimates
of mortality due topre-natal conditions were obtained bycalculating the attributable fraction
of neonatal mortality due to IUGR, then multiplyingthis value by the estimated attributable
fraction of IUGR due tolow maternal pre-pregnancy BMI. The attributable fractions were
based on prevalence and risk estimates from published and unpublished sources. The
relationship between maternalunderweight status and neonatal mortality was estimated by
deriving the proportion of IUGR attributable to poor maternal pre-pregnancy
anthropometric status and theproportion of neonatal mortality attributed to IUGR. The
equation thus has fourcomponents: the proportion of IUGR births among live births; an
estimate of infantmortality risk associated with IUGR; prevalence of underweight status
among women of reproductive age; and anestimate of the risk of IUGRassociated.
21. Anderson, J.R., R.W. Herdt and G.M. Scobie, 1988. %
j $+
World Bank, Washington DCwith pre-pregnantunderweight status (Fishman et al.,
2004).

c
c
c c
More details on the quantitative relationshipsare provided in Section 3. A recent analysis

that compared different causes of mortality and morbidity showed that maternal and child
under nutrition is the single leading cause of health loss worldwide (Ezzati et al., 2002).22
This alarming statistic underscores the huge potential for decreasing global ill-health if
malnutrition were to be reduced or even eliminated.

((+")*6$+"+3(+$)+
While episodes of severe hunger such as famines receive considerable press coverage and
attractmuch public attention, chronic hunger and malnutrition is considerably more
prevalent in developingcountries. It is estimated that at least 12 million low-birth-weight
births occur per year and that around162 million pre-school children and almost a billion
people of all ages are malnourished. In poorlynourished populations, reductions in hunger
and improved nutrition convey considerable productivitygains as well as saving resources
that otherwise would be used for the care of malnourished people whoare more susceptible
to infectious diseases and premature mortality. While reducing hunger andmalnutrition is
often justified on intrinsicgrounds, it is these potential gains in productivity andreductions
in economic costs thatprovide the focus of our challenge paper. 23Poverty, hunger and
malnutrition are linked. Strauss and Thomas (1995, 1998) and Hoddinott,Skoufias and
Washburn (2000) document the empirical literature relating dimensions of access andintakes
of calories to household consumption levels. A reasonable reading of these studies suggests
anincome-calorie elasticity or around 0.2 to 0.3, though careful studies have also found
estimates bothhigher and lower than this range. Behrman and Rosenzweig (2004) report that
cross-country variation inGDP per capital in PPP terms is inversely related to the percentage
of low birth weight (LBW, <2.5 kg,)births among all births and is consistent with almost half
of the variation inthe percentage of births thatare LBW across countries.
22. Angrist, J., E. Bettinger, E. Bloom, E. King and M. Kremer, 2002: Vouchers for Private
Schooling in Colombia: Evidence from a Randomized Natural Experiment,
)

92(5): 1535-59.
23. Ashworth, A., 1998: Effects of Intrauterine Growth Retardation on Mortality and
Morbidity in Infants inYoung Children, )
"

%$
4(Suppl): S34-
42.

c
c
c c
Haddad, et al. (2003) estimate that the cross-country elasticity of pre-schoolunderweight

rates with respect to per capita income for 1980-96 is -0.5, virtually the same as the mean
forthe elasticity from 12 household data sets.24These relationships have two important
implications: thatnutritional objective such as the Millennium D evelopment Goal of halving
the prevalence ofunderweight children by 2015 is unlikely to be met through income growth
alone and that successfulefforts to reduce most forms of malnutrition are likely to have
incidences of benefits concentratedrelatively among the poor. These implications motivate,
in part, our choice of the following opportunities:
Opportunity 1 ² Reducing the Prevalence of Low Birth weight
Opportunity 2 ² Infant and Child Nutrition and Exclusive Breastfeeding Promotion
Opportunity 3 ² Reducing the Prevalence of Iron Deficiency Anemia and Vitamin A,
Iodine andZinc Deficiencies.25
Opportunity 4 ² Investment in Technology in Developing Country Agriculture
We begin by setting the stage, discussing the nature and measurement of hunger and
malnutrition,he current levels and trends in the geographical distribution among
developingcountries of someimportant types of hunger and the nature of the benefits from
reduced malnutrition both in terms ofproductivity and in terms of direct resource use. This
is essential to avoid repetition because many of themeasurement issues, including those
pertaining to impacts of improved nutrition over the lifecycle, aresomewhat parallel among
the various challenges and opportunities.
Section 3 outlines a generalframework for considering these opportunities. 26 For each
opportunity, we discuss: (i) the definition anddescription of the opportunity, (ii) how this
opportunity partially solves the challenge, and (iii) economicestimates of the benefits and
costs and how they relate to distributional and efficiency motives forpolicies. Our conclusion
summarizes these opportunities, noting that potentially
24. Ashworth, A., S. Morris and P. Lira, 1997: Postnatal growth patterns of full-term low
birth weight infantsin North east Brazil are related to socioeconomic status, "

%

, 127: 195-1956.
25. Aylward, G. 2003: Cognitive Function in Preterm Infants. No Simple Answers, "

%

, 289(6): 747-750
26. Aylward, G., S. Pfeiffer, A. Wright and S. Verhulst, 1989: Outcome Studies of Low Birth
weight InfantsPublished in the Last Decade: A Meta-Analysis, "

%, , 115:515-
520.
c
c
c c
there are considerablegains in the sense of benefit-cost ratios exceeding one or relatively
high internal rates of return toinvesting in some programs or policies to reduce hunger and
malnutrition, particularly those directedtowards increasing micronutrients in populations
with high prevalence of micronutrient deficiencies ² inaddition to the intrinsic welfare gains
to the individuals who would be effected directly by reduced hungerand malnutrition.
Such investments may be relatively easily justified on an ti-poverty grounds because the
poor tend to be relatively malnourished. 27 There also are some plausible efficiency grounds
for suchinterventions due to the role that malnourishment plays, for example, in the spread
of contagious diseasesbut the available estimates do not permit very satisfactory
identification of social versus private rates ofreturns as would be required to assess the
efficiency motive for subsidies. 28
+(51")$+3)((+"+3 77)$+ )*

6$+"+3(+$)+- $+3$&+)"7(
'$)++3,)+(5+*
This section provides background material for understanding the opportunities in Section 3.
Itdiscusses the nature and measurement of hunger and malnutrition, current levels and
trends in thegeographical distribution of dimensions of hunger and malnutrition, and the
nature of the benefits fromreduced hunger and malnutrition, which are parallel over the life
cycle for several of the opportunities.

+$+3&$&+)*$+"+3&(+$)+
Hunger can be defined as Á condition, in which people lack the basic food intake to
providethem with the energy and nutrients for fully productive livesµ (Hunger Task Force,
2003, p. 33). It is measured in terms of availability, access or intake of calories relative to
requirements that vary principally by age, sex and activities.
27.Ballard, C., J. Shoven, and J. Whalley, 1985: General Equilibrium Computations of the
Marginal Welfare Costs of Taxes in the United States, )

å4(1): 128-
38.
28. Barker, D.J.P., ed., 1992: Fetal and Infant Origins of Adult Disease: Papers written by The
Medical Research Council Environmental Epidemiology Unit, University of Southampton,
London: British Medical Journal.
c
c
c c
Nutrients provided by food combine with other factors, including the health state of the
person consuming the food, to produce ńutritional status.µ Indicators of nutritional status
are measurements of body size, body composition, or body function reflecting single or
multiple nutrient deficiencies. Table 1 describes measures of hunger and under nutrition
important to this Challenge Paper.2 The most widely-cited data on the number of persons
considered hungry come from the United Nations Food and Agriculture Organizat ion
(FAO). On an ongoing basis, FAO constructs estimates of mean per capita dietary energy
supply (production + stocks - post-harvest losses + commercial imports + food aid -
exports). Assumptions regarding the distribution of this supply are made based on data
onincome distribution, the distribution of consumption or, in some cases inferences based on
infant mortality (Naiken, 2002). The constructed distribution is compared against minimum
per capita energy requirements (Weisell, 2002) and from this, the proportion of persons
whose access to food is below these requirements is estimated. FAO calls this the prevalence
of undernourishment. 29
Criticisms of this approach are widespread. First, there are serious concerns about the
quality of the underlying data on food supply (Devereux and Hoddinott, 1999). Second, the
absence of good data on the distribution of food consumption mean that estimates of the
prevalence of hunger are highly sensitive to changes in the shape of the distribution around
the minimum requirements threshold. Third, Aduayom and Smith (2002) show that in many
cases the FAO approach significantly understates hunger prevalence when compared to
those derived from household consumption surveys. 30
Despite these valid concerns, the FAO approach provides the only data available on a global
basis over a relatively long time period. FAO (2003) estimates that over the last decade, the
number of people undernourished in the developing world declined slightly from 816 to
798 million while population increased from 4050 to 4712 million persons. Overall, the
proportion of persons undernourished fell from 20 to 17% between 1990-92 to 1999-2001.
29. Barker, D.J.P., 1998:

% - , Churchill Livingstone,
Edinburgh, London, New York.
30.Barrera, A., 1990: The Role of Maternal Schooling and Its Interaction with Public Health
Programs in Child Health Production, "

%&
)

32(1): 69-92.

c
c
c c
The hungry are found predominantly in Asia and the Pacific (505 million) and sub-Saharan
Africa (198 million); these two regions account for nearly 90% of the world·s hungry.
However, these two regions exhibit different trends. 31 In Asia, both the number and
prevalence of undernourishment fell during the 1990s. The fall in the number of
undernourished is almost entirely attributable to a fall in the number of undernourished in
China; elsewhere the number of undernourished stayed r elatively constant while population
grew, leading to a decline in prevalence. In Africa, the number of malnourished increased
with prevalence rising in some countries and falling in others. Despite these shifts, in the
near future over twice as many of the hungry will be in Asia than in Africa. The distribution
of the hungry countries and by socio-economic groups is even less well
documented. 32
Preliminary work by the Hunger Task Force (2003) suggests that on a global basis:
Approximately 50% of the hungry are in farm households, mainly in higher -risk
productionenvironments;
Approximately 25% are the rural landless, mainly in higher-potential agricultural regions;
Approximately 22% are urban; and
Approximately 8% are directly resource-dependent (i.e. pastoralists, fishers, forest-based).
Both hunger and malnutrition reflect the interaction of purposive actions of individuals
givenpreferences and constraints together with biological processes. In behavioral models,
an individual·s nutritional status often is treated as an argument in the welfare function of
individuals or the households in which they reside (Behrman and Deolalikar 1988; Strauss
and Thomas 1995), a reflection of the intrinsic value placed on nutritional status. Typically,
welfare is assumed to increase as nutritional status improves, but possibly at a diminishing
rate and increases in certain measures of nutritional status, such as body mass, may be
associated with reductions in welfare beyond a certain point. In allocating resources,
31. Behrman, J. R. and A. Deolalikar, 1988: Health and Nutrition. In H. B. Chenery and T.N.
Srinivasan, eds.,

)

&
, Vol. 1, 631-711, North Holland
Publishing Co., Amsterdam.
32. Behrman, J. R. and A. Deolalikar, 1989: Wages and Labor Supply in Rural India: The Role
of Health, Nutrition and Seasonality. In D. Sahn, ed., $

%

.

a
, 107-118, The Johns Hopkins University Press,

Baltimore, MD.
c
c
c c
household decision makers take into account the extent to which these investments will
make both their children and themselves better-off in the future as well as currently. These
allocations are constrained in several ways. There are resource constraints reflecting income
(itself an outcome becausenutritional status can affect productivity) and time available as
well as prices faced by households. There is also a constraint arising from the production
process for health outcomes, including nutritional status. This constraint links nutrient
intakes ² the physical consumption of macronutrients (calories and protein) and
micronutrients (minerals and vitamins) ² as well as time devoted to the production of health
and nutrition, the individual·s genetic make-up and knowledge and skill regarding the
combination of these inputs to produce nutritional status. There are interdependencies in the
production of nutritional status and other dimensions of health; for example, malaria limits
hemoglobin formation.33
Many poor nutritional outcomes begin

. A number of maternal factors have been
shown to be significant determinants of intrauterine growth retardation (IUGR), the
characterization of a newborn that does not attain their growth potential. Most important
are mother·s stature (reflecting her own poor nutritional status during childhood), her
nutritional status prior to conception as measured by her weight and micro-nutrient status,
and her weight gain during pregnancy. 34
Diarrheal disease, intestinal parasites, and respiratory infections may also lead to IUGR and
where endemic (such as in sub-Saharan Africa), malaria is a major determinant. In
developed countries, smoking is also a significant contributor to IUGR. IUGR is measured as
the prevalence of newborns below the 10th percentile for weight given gestational age
(ACC/SCN, 2000b). Because gestational age is rarely known, IUGR is often proxied byLBW.
As of 2000, it is estimated that 16% of newborns, or 11.7 million children have
LBW(ACC/SCN, 2000b, ACC/SCN, 2003).3 LBW is especially prevalent in south Asia
where it is estimated that 30 per cent of children have birth weights below 2500g
(ACC/SCN, 2003).
33. Behrman, J., A. Foster and M. Rosenzweig, 1997: The Dynamics of Agricultural
Production and the Calorie-Income Relationship: Evidence from Pakistan, "

%
)

åå(1): 187-207.
34. Behrman, J. R. and J. Hoddinott, 2002: Program Evaluation with Unobserved
Heterogeneity and Selective Implementation: The Mexican ,
Impact on Child
Nutrition, University of Pennsylvania, Philadelphia, mimeo.
c
c
c c
In pre-school and school-age children, nutritional status is often assessed in terms of

anthropometry.35 ´The basic principle of anthropometry is that prolonged or severe nutrient
depletion eventually leads to retardation of linear (skeletal) growth in children and to loss
of, or failure to accumulate, muscle mass and fat in both children and adultsµ (Morris, 2001,
p.12). A particularly useful measure is height-for-age as this reflects the cumulative impact
of events affecting nutritional status that result in stunting. As of 2000, it is estimated that
162 million children ² roughly one child in three ² are stunted (ACC/SCN, 2003).
While stunting prevalence are highest in South Asia and sub -Saharan Africa, in South Asia,
numbers and prevalence have been declining since 1990 whereas in sub-Saharan Africa,
prevalence has remained largely unchanged and numbers have increased.Multiple factors
contribute to poor anthropometric status in children. One is LBW; a number of studies sh ow
a correlation between LBW and subsequent stature though, in the absence of any subsequent
intervention, not between LBW and growth (Ashworth, Morris and Lira 1997; Hoddinott
and Kinsey.
Growth rates are highest in infancy, thus adverse factors have a gr eater potential for causing
retardation at this time. Younger children have higher nutritional requirements per kilogram
of body weight and are also more susceptible to infections. They are also less able to make
their needs known and are more vulnerable to the effects of poor care practices such as the
failure to introduce safe weaning foods in adequate quantities. Evidence from numerous
studies clearly indicates that the immediate causes of growth faltering are poor diets and
infection (primarily gastrointestinal) and that these are interactive (Chen 1983; NAS 1989).36
For these reasons, almost all the growth retardation observed in developing countries has its
origins in the first two to three years of life (Martorell 1995). While much of the work on
hunger and nutrition is cross-sectional, an increasing body of knowledge indicates that
many nutritional outcomes are the consequence of cumulative lifecycle processes.
35. Behrman, J. R., J. Hoddinott, J. A. Maluccio, Agnes Quisumbing, R. Martorell and Aryeh
D. Stein, The Impact of Experimental Nutritional Interventions on Education into Adulthood
in Rural Guatemala: Preliminary Longitudinal Analysis, 2003: University of Pennsylvania,
IFPRI, Emory Philadelphia - Washington-Atlanta, processed.
36. Behrman, J., and J. Knowles, 1998a: Population and Reproductive Health: An Economic
Framework for Policy Evaluation, ,

&
24*4): 697-738.

c
c
c c
Specifically, a growing body of evidence indicates that growth lost in early years is, at best,
only partially regained during childhood and adolescence, particularl y when children
remain in poor environments (Martorell, et al. 1994).
Martorell (1995, 1999), Martorell, Khan and Schroeder (1994) and Simondon, et alÊ (1998) all
find that stature at age three is strongly correlated with attained body size at adulthood in
Guatemala and Senegal. Hoddinott and Kinsey (2001) and Alderman, Hoddinott and Kinsey
(2003) find that children who were initially aged 12-24 months in the aftermath of droughts
in rural Zimbabwe in 1994/95 and 1982-4 respectively were malnourished relative to
comparable children not exposed to this drought. However, older children did not suffer
such consequences; this is consistent with evidence that child development has ´sensitiveµ
periods where development is more receptive to influence and that during such periods,
some shocksmay be reversible while others are not. 37
Severe malnutrition in early childhood often leads to deficits in cognitive development
(Grantham-McGregor, Fernald and Sethuraman 1999, Pollitt 1990). 38
Though many studies from developed countries fail to show difference in developmental
levels for children with LBW, there are few longitudinal studies from developing counties
from which to generalize (Hack 1998). Moreover, recent studies indicate that the relationship
between birth weight and cognitive function carries into the range of normal weights even
in developed countries (Richards, et al. 2001; Matte, et al. 2001).
Reduced breastfeeding ² an effect of LBW as well as a common cause of childhood
malnutrition - also has well documented influences on cognitive development, even in
developed countries (Grantham-McGregor, Fernald and Sethuraman 1999a). Malnourished
children score more poorly on tests of cognitive function, have poorer psychomotor
development and fine motor skills, have lower activity levels, interact less frequently in their
environments and fail to acquire skills at normal rates (Grantham-McGregor, et al.1997,
1999; Johnston, et al. 1987; Lasky, et al. 1981).
37.Bobonis, Gustavo, E. Miguel, and Charu Sharma. 2003: Child Nutrition and Education: A
Randomized Evaluation in India, U.C. Berkeley, unpublished working paper.
38. Boissiere, M., J. Knight and R. Sabot, 1985: Earnings, Schooling, Ability and Cognitive
Skills, )

75: 1016-30.
c
c
c c
Controlled experiments with animals suggests that malnutrition results in irreversible

damage to brain development such as that associated with the insulation of neural fibers
(Yaqub 2002). This is in keeping with the prevailing view that very young childr en are most
vulnerable to impaired cognitive development.4 Micro-nutrient deficiencies, particularly
iodine and iron, are also stronglyimplicated in impaired cognitive development. Iodine
deficiency adversely affects development of the central nervous system. 39A meta-analysis
indicates that individuals with an iodinedeficiency had, on average, 13.5 points lower IQs
than comparison groups (Grantham-McGregor, Fernald and Sethuraman 1999b). ACC/SCN
(2003) reports that globallyapproximately 2 billion people are affected by iodine deficiency
including285million childrenaged 6 to 12 years.
More than 40% of children ages 0-4 in developing countries suffer from anemia (ACC/SCN
2000a); further anemia in school-age children may also affect schooling whether or not there
had been earlier impaired brain development. Under nutrition, particularly fetal under
nutrition at critical periods, may result in permanent changes in body structure and
metabolism. Even if there are not subsequent nutritional insults, these changes can lead to
increased probabilities of chronic non-infectious diseases later in life. The hypothesis that
fetal malnutrition has far ranging consequences for adult health is bolstered by studies that
track LBW infants into their adult years and document i ncreased susceptibility to coronary
heart disease, noninsulindependent diabetes, high blood pressure, obstructive lung disease,
high blood cholesterol and renal damage (Barker 1998). For example, while the various
studies on the impact of the Dutch famine indicate few long-term consequences on young
adults, more recent evidence shows that children whose mothers were starved in early
pregnancy have higher rates of obesity and of heart disease as adults (Roseboom, et al. 2001).
In contrast, children of mothers deprived in later pregnancy ² the group most likely to be of
LBW ² had a greater risk of diabetes (Ravelli 1998).Adequate iron intake is also necessary for
brain development. 40
The nutritional status of adults reflects in substantial part their nutritio nal experience
sinceconception with, as noted, a number of possible long-run effects of early nutritional
insults.
39. Bouis, H., 2002a: Plant breeding: A new tool for fighting micronutrient malnutrition,
"

%
132: 491S-494S.

c
c
c c
40. Bouis, H., 2002b: Appendix to: Productivity losses from iron deficiency: The economic
case, by Joseph Hunt, "

%
132: 794S-801S.
But, in addition to such longer-run effects, there also may be important consequences of
adult diets, for example, low energy or iron intakes or chronic diseases related to obesity,
hypertension, and high cholesterol. 41 Lastly, malnutrition may have long-term consequences
through theintergenerational transmission of poor nutrition and anthropometric status.
Recall that stature by age three is strongly correlated with attained body size at adulthood.
Taller women experience fewer complications during childbirth, typically have children
with higher birth weights and experience lower risks of child and maternal mortality
(Ramakrishnan, et al. 1999; World Bank 1993)42.
+$)*'+**)&3$3&(+$)+
One exception is provided by Glewwe and King (2001) who find that malnutrition in the
second year of life had a larger impact on the IQs of Philippine school children than that in
earlier periods.43 This may reflect that with weaning risks increase.However, Behrman and
Rosenzweig (2004) find that intergenerational birth weight effects are primarily genetic, not
due to better nutrition in the womb, based on their analysis of identical twins in the United
States.44
It is not clear whether this result generalizes to developing countries because there may be
important compensating investments for LBW in developed societies that are not common
in developing countries (and that may be reflected, for example, in the evidence noted
suggesting stronger effects on cognitive development in the latter). We now turn to micro
evidence about productivity impacts of improved nutrition in developing countries ² from
41. Bourdillon, M., P. Hebinck, J. Hoddinott, B. Kinsey, J. Marondo, N. Mudege, and T.

Owens, 2003: Assessing the impact of HYV maize in resettlement areas of Zimbabwe, FCND
Discussion paper No 161, International Food Policy Research Institute, Washington D.C.
42. Boyle, M. G. Torrance, J. Sinclair and S. Horwood, 1983: Economic Evaluation of
Neonatal Intensive Care of Very Low Birth-Weight Infants, ) "

% ,
:1330-37.
43. Brabin, B., M. Hakimi, and D. Pelletier, 2001: An analysis of anemia and pregnancy-
related maternal mortality, "

%
, 131(2S-2): 604S-614S.

c
c
c c
44. Brown K., J. Peerson, J. Rivera, and L. Allen, 2002: Effect of supplemental zinc on the
growth and serum zinc concentrations of pre-pubertal children: a meta-analysis of
randomized controlled trials, "

%$
75:1062²71.
conception through infancy and childhood and into adolescence and
adulthood. 45Thesemany channels through which these gains may operate are grouped as
follows: saving of resources that are currently directed to dealing with diseases and other
problems related to malnutrition; direct gains arising from improvements in physical stature
and strength as well as improved micronutrient status; and indirect gains arising from links
between nutritional status and schooling, nutritional status and cognitivedevelopment and
subsequent links between schooling, cognitive ability and adult productivities.

)$%+"One significant cost of malnutrition is higher mortality. The probabilityof
infant mortality is estimated to be significantly higher for LBW than for non -LBW infants.
Conley, Strully and Bennett (2003) conclude that intra-uterine resource competition ² and,
by inference, nutrition ² explain a substantial portion of excess mortality of LBW children in
the United States. In their study, an additional pound at birth led to a 14% decrease in
mortality in the period between 28 days and one year for both fraternal and identical twins.
In contrast, the risk of death in the first 28 days was elevated 27% for each pound difference
in weight for fraternal twins compared to only 11% for identical twins, implying a large role
for genetic factors. Ashworth (1998) reviews 12 data sets including two from India and one
from Guatemala, and concludes that the risk of neonatal death for term infants 2000-2499
gramsat birth is four times that for infants 2500-2999 grams and 10 times that of infants 3000-
3499 grams. Relative risks of post-neonatal mortality for LBW compared to the two
respective groups were two and four times as large. These risk ratios translate into fairly
large differences in mortality rates given the relatively high mortality rates in many
developing countries (see the discussion of Opportunity 1). When the impacts of poor pre -
schooled nutrition are added to the effects of LBW, Pelletier, et al. (1995) venture the widely -
cited estimate that 56% of child deaths in developing countries at attributable to the
potentiating effects of malnutrition (83% of this, due to the more prevalent mild to moderate
malnutrition rather than the severe cases most commonly monitored). 46
More recently, WHO (2002) has claimed that malnutrition contributes to 3.4 million child
deaths in 2000 (60% of child deaths).7

c
c
c c
45. Byerlee, D., 1993: Technical change and returns to wheat breeding research in Pakistan·s
Punjab in the post -Green Revolution period, , &
, 32(spring): 69-86.
46. Cameron, N. 2001: Catch-up Growth Increase Risk Factors for Obesity in Urban Children
in South Africa,
"

% 25(Suppl.): S48.
There are aggregate or macro alternatives, such as to define benefits in terms of an
investment·s impact on economic growth, typically measured in terms of growth in GNP per
capita. This approach was used in the pioneering study by Coale and Hoover (1958) of the
economic benefits of fertility reduction and more recently in combination with other
methods to estimate the economic benefits of a broad strategy to improve health in
developing countries (Commission on Macroeconomics and Health 2001). The latter·s
estimates of the relationship between cross-country economic growth rates, indicators of
population health and a set of additional explanatory variables indicate that each10%
improvement in average life expectancy at birth is associated with an increase inthe rate of
economic growth of 0.3 to 0.4 percentage points per year. However, the associations found
in cross-national analysis are unlikely to represent unbiased estimates of the causal effects of
investments due to non-trivial omitted variable and endogeneity bias. For example,
Behrman and Rosenzweig (2004) explore the relation between birth weight and productivity
across countries: aggregate estimates indicate an association with over 40% of the variance in
product per worker across countries, but estimates that control for micro endowments
suggest that lessthan 1% of cross-country differences can be attributed to differences in birth
weight distributions. Further, most of the available cross-national data are not sufficiently
disaggregated to disentangle the effects of the types of investments being considered by the
Copenhagen Consensus from similar broader investments in the popu lation.47
7 Pelletier and Frongillo (2003), using data on changes in national malnutrition rates and
mortality to get a different perspective on this association, also find an association of
mortality and malnutrition. However, these associations do not c ontrol for changes in
infrastructure or income that may both affect mortality directly as well as influence nutrition
nor can they indicate a counterfactual of the impact of improved nutrition on expected
mortality. Guilkeyand Riphahn (1998) use longitudin al data on Filipino children with
controls for the endogeneity of nutrition and other health care choices. 48

c
c
c c
47. Castleman, T., et al., 2003: Food and nutrition implications of antiretroviral therapy in
resource limited settings, FANTA Technical Note No. 7, Washington DC.
48. Caulfield, L., N. Zavaleta, A. Shankar, and M. Merialdi, 1998: Potential contribution of
maternal zinc supplementation during pregnancy to maternal and child survival,
"

%$
499S-508.
Their simulations indicate that children with two months without weight gain in the first
year of life (about 10 percent of their sample) have the risk of mortality elevated by
50%.Similarly, the scenarios show that if a mother is unable or unwilling to adopt standard
recommendations on breastfeeding the hazard of child The availability of experimental
evidence on the use of micronutrient supplements provides unambiguous evidence on
therelationship of mortality and vitamin intakes in many environments including ones that
showfew clinical symptoms of deficiencies. The potential to reduce child deaths by
distributingvitamin A on a semiannual basis isparticularlydramatic; meta-analysis of field
trials indicatethat such provision of vitamin A can reduce overall child mortality by 25-35%
(Beaton, et al. 1993)Amongst adults, anemia is a particular concern for the health of women
of child bearing age not only because of elevated risk of adverse birth outcomes but also
because the risk of maternal death is substantially elevated for anemic women; over a fifth of
maternal deaths are associated with anemia (Ross and Thomas 1996; Brabin, Hakimi and
Pelletier 2001).
Beyond the issue of increased mortality, malnutrition increases the risk of illnesses that
impair the welfare of survivors. This relationship between nutrition and both infections and
chronic diseases can be traced through different parts of the life cycle. Children with LBWs ²
reflecting a range of causes, not all of which are due to dietary deficiencies ² stay longer in
hospitals in circumstances where births occur in such settings and have higher risks of
subsequent hospitalization (Vitoria, et al. 1999). 49In addition, they use outpatient services
more frequently than do children with normal birth weights. For young chil dren, in general,
malnutrition, including micronutrient deficiencies, leads to a vicious cycle with impaired
immunity leading to infection with attendant loss of appetite and increased catabolism and,
thus, an increased likelihood of additional malnutrition. Increased morbidity has direct
resource costs in terms of health care services as well as lost employment or schooling for
the caregivers. The magnitudes of these costs differ according to the medical system,
markets, and policies of a country. In developed countries the costs for the survivors can be
substantial.
c
c
c c
49. Cawley, J., J. Heckman and E. Vytlacil, 2001: Three Observations on Wages and
Measured Cognitive Ability,
)

, : 419-442. Ceesay S., A. Prentice, T. Cole, R.
Ford, E. Poskitt, L. Weaver, and R. Whitehead, 1997: Effects on birth weight and perinatal
mortality of maternal dietary supplements in rural Gambia: 5 year randomized controlled
trial, "
315: 786-790.
Similarly,75% of the $5.5-6 billion of excess costs due to LBW in the United S tates estimated
by Lewit, et al (1995) is due to the costs of health care in infancy. A further 10% of these costs
are attributed to higher expenditures for spec ial education as well as increased grade
repetition. Such expenditures for special education or for social services are substantial in
developedcountries (Petrou, et al. 2001). While these costs may be far less in low-income
countries where, for example, the majority of births occur outside a clinical setting, these
lowermedicalcostsassociated with LBW come at the expense of higher mortality.In the
absence of an educational system that can recognize and accommodate the individual needs
of students, moreover, some of these costs are not incurred during childhood but rather in
the form of reduced productivity in adulthood. The evidence for the fetal origins of chronic
diseases described above is still being assessed. The fact that some consequences may not be
observed until the affected individuals reach middle age is an important consideration for
interpreting the range of evidence being assembled. There are few longitudinal studies that
follow cohorts this far and extrapolation from shorter panels or from cohorts with different
life histories are problematic. In addition, there are at least two other explanations for the
association between LBW and adult diseases. LBW may be an indicator of poor
socioeconomic status.
Low SES may have a causal impact on adult disease probabilities via other variables such as
poor nutrition later in life or higher rates of smoking. If so, LBW may only be a correlate and
not a causal variable. 50 A different possibility is that LBW may be due to a genetic
predisposition to insulin resistance. This would tend to account for a higher pre-disposition
for adult diabetes and coronary heart diseases that reflects genetics rather than aspects of the
uterine environment that may be influenced by medical and mortality increases markedly.
Care has to be taken, however, in interpreting the last association as causal becausemothers
may be less likely to be able to breastfeed infants who are at high risk.
8 As with associations of child mortality and nutrition, it is difficult to prove causality with
these associations. Nutritional interventions.9 Finally, even if there are the effects proposed
in the fetal origins hypothesis, due to the long lags, the present discounted value of

c
c
c c
improvements due to prenatal interventions to offset them is not likely to be very large
(Alderman and Behrman 2003; also see Opportunity 1).
50. Ching, P., M. Birmingham, T. Goodman, R. Sutter and B. Loevinsohn, 2000: Childhood
Mortality Impact and Costs of Integrating Vitamin A Supplementation Into Immunization
Campaigns, "

%, , 90(10): 1526-1529.
Direct links between nutrition and physicalproductivity: There is considerable evidence of
direct links between nutrition and productivity. Behrman (1993), Behrman and Deolalikar
(1989), Deolalikar (1988), Foster and Rosenzweig (1993), Glick and Sahn (1997), Haddad and
Bouis (1991), Schultz (1996), Strauss and Thomas (1998) and Thomas and Strauss (1997) all
find that after controlling for a variety ofcharacteristics, that lower adult height ² as
described above, a consequence in part of poornutrition in childhood, is associated with
reduced earnings as an adult. Thomas and Strauss (1997) estimate the direct impact of adult
height on wages for urban Brazil. 51 While the elasticity varies somewhat according to gender
and specification, for both men and women who work in the market sector a 1% increase in
height leads to a 2-2.4% increase in wages or earnings.10 While their study is relatively
sophisticated in the methodology used to account for labor selectivity and joint
determination of health, this result is similar to others reported in the literature. Low energy
intakes can reduce productivity creating a vicious circle in which poor workers are unable to
generate sufficient income to obtain sufficient calories to be productive. This relationship,
sometimes dubbed the efficiency wage hypothesis, has been the object of considered
theoretical work since Leibenstein (1957). Strauss and Thomas· (1998) review of the
empirical literature notes that efforts to test this relationship empirically have been dogged
by a number of problems: unobservable heterogeneity, measurement error, and
observability issues.
They note, Ít is not obvious how to interpret a result that additional calories are associated
with higher productivity if higher productivity workers are stronger and consume more
caloriesµ (p. 806). While individual fixed effects specifications can addressunobservable
fixed heterogeneity, they are especially susceptible to problems derived from measurement
error. Thomas and Strauss (1997) lessen the latter problem by drawing on micro data in
which caloric intakes were measured directly over a seven day period; in the Brazil data,
they find that wages of workers in urban areas were positively and significantly affected by
calories at low intake levels.

c
c
c c
51. Coale, A. and E. Hoover, 1958: ,

+
)

&

$
$
% /,
, Princeton University Press, Princeton, NJ.
Cobra, C., K.R. Muhilal D. Diet Rustama S. Suwardi D. Permaesih S.M.
Muherdiyantiningsih and R. Semba, 1997: Infant Survival Is Improved by Oral Iodine
Supplementation, "

%
, 127: 574-578.
Foster and Rosenzweig (1993a, 1994) find that calorie intakes have a significant effect on
piece rates (but not time rates) in the Philippines where some workers engage in both time
and piece rate activities so it is possible to control for unobserved individual heterogeneities.
Micronutrient status also has important productivity effects. Vitamin A deficiency can cause
blindness with obvious consequences for productivity. Anemia is associated with reduced
productivity both in cross-sectional data and in randomized interventions (Thomas, et al.,
2004,Li, et al. 1994; Basta, Karyadi and Scrimshaw 1979). The magnitude may depend on the
nature of the task. For example, piece work may have greater incentives for effort while
heavy physical labor may show greater increases in productivity, though anemia is
nevertheless a factor in productivity with relatively light work (Horton and Ross 2003).52
Poorly nourished children tend to start school later, progress through school less rapidly,
have lower schooling attainment, and perform less well on cognitive achievement tests
when older, including into adulthood. These associations appear to reflect significant and
substantial effects in poor populations even when 9 An additional aspect of the hypothesis
of subsequent costs stemming from biological adaptation to deprivation
has a
bearing on the estimation of the consequences of LBW.
That is, the implications will be different if the consequences are a direct result of the
deprivation compared to the possibility that they only manifest themselves ifthe deprivation
is followed by relative abundance (Lucas, Fewtrell and Cole 1999, Cameron 2001). 10 From a
different perspective, Margo and Steckel (1982) found that the value of an American slave
prior to the United States· Civil War fell by roughly 1.5 percent for every reduction in height
of one inch. Statistical methods such as instrumental variables are used to control for the
behavioral determinants of pre-school malnutrition. In productivity terms, the magnitudes
of these effects are likely to be substantial, easily exceeding the effects of height on
productivity even if the indirect effect of height on wages mediated by the relationship
between height and schooling in included.11 There are at least three broad means by which
nutrition can affect education.
c
c
c c
52. DeLong. G., P. Leslie, S-H. Wang, X-M. Jiang, M-L. Zhang, M. Abdul Rakeman, J-Y.
Jiang, T. Ma and X-Y Cao, 1997: Effect on infant mortality of iodination of irrigation water in
a severely iodine-deficient area of China, , 350: 771-773.
First, malnourished children may receive less education. This may be because their
caregivers seek toinvest less in their education, because schools use physical size as a rough
indicator of school readiness or because malnourished children may have higher rates of
morbidity and thusgreater rates of absenteeism from school and learn less while in school.
53 While delayed entry, the second way by which nutrition may influence schooling, does not
necessarily lead to less completed schooling ² although under standard models of the
returns to schooling this would be an expected consequence of delayed enrollment if the
opportunity cost of schooling increases with age ² late enrollment leads to lower expected
lifetime earnings. In order to maintain total years of schooling with delayed entry, an
individual has to enter the work force when older. As Glewwe and Jacoby (1995) illustrate,
for each year of delay in entry to primary school in Ghana a child in their study loses 3
percent of lifetime wealth. The third pathway from malnutrition to educational outcomes is
via the capacity to learn, a direct consequence of the impact of poor nutrition on cognitive
development described above. 54
Additionally, a hungry child may be less likely to pay attention in school and, thus, learn
less even if he or she has no long-term impairment of intellectual ability.12 These three
pathways clearly interact; a child with reduced ability to learn will likely spend less time in
school as well as learn less while in class. While there are many studies that document
associations between nutrition and schooling (see Pollitt 1990 and Behrman 1996 for
reviews), there are far fewer studies that persuasively portray the causal impact of child
health and nutrition on school performance. Many of the observable factors that affect
nutrition, such as family assets and parental education, are also ones that affect education.
Similarly, unobservable attitudes aboutinvestment in children and intra -family equity
influence heath provision and schooling decisions in a complex manner. 55Four recent studies
represent the most complete efforts at distinguishing the distinct contributory role of
53. Deolalikar, A., 1988: Nutrition and Labor Productivity in Agriculture: Estimates for Rural
South India,
%)

70(3): 406-413.

c
c
c c
54. De Onis, M., M. Blossner, and J. Villar, 1998: Levels and Patterns of Intrauterine Growth
Retardation in Developing Countries, )
"

%$
, 52(S1): S5-S15.
55. Deolalikar, A., 1988: Nutrition and Labor Productivity in Agricultu re: Estimates for Rural
South India, !
%)

70(3): 406-413. De Onis, M., M. Blossner, and J.
Villar, 1998: Levels and Patterns of Intrauterine Growth Retardation in Developing
Countries, )
"

%$
, 52(S1): S5-S15.
nutrition on education from associations.Glewwe, Jacoby, and King (2001) track children
from birth through primary school and find that better nourished children both start school
earlier and repeat fewer grades. A 0.6 standard deviation increase in the stature of
malnourished children would increase completed schooling by nearly 12 months. Using
longitudinal data from rural Pakistan where school initiation is much lower, Alderman, et
al. (2001) find that malnutrition decreases the probability of ever attending school,
particularly for girls. An improvement of 0.5 standard deviations in nutrition would
increase school initiation by 4% for boys but 19% for girls. As the average girl (boy) in the
villages studied who begins school competes 6.3 (7.6) years of schooling, improvements in
nutrition would have a significant effect on schooling attainment. Alderman, Hoddinott, and
Kinsey (2003) track a cohort of Zimbabweans over two decades finding that
delayed
school initiations and fewer grades completed for those children malnourished as children. 56
Extrapolating beyond the drought shocks used for identification, the study concludes that
had the median pre-school child in the sample achieved the stature of a median child in a
developed country, by adolescence she would be 4.6 centimeters taller, had completed an
additional 0.7 grades of schooling as Strauss and Thomas (1998) point out that an illiterate
man would need to be 30 cm taller than his literate coworker to have the same expected
wage.
A few studies have attempted to investigate the tie between hunger and classroom
performance using experimental designs. Available results, however, are not conclusive
regarding long-term consequences, perhaps, in part, because controlled studies are
hampered by difficulties in running experiments for an appreciable duration as well as the
difficulty of encouraging parents to conform to the protocols of research design and the
inability to use aplacebo. Moreover, as shown in Grantham-McGregor, Chang and Walker
(1997), while feeding children may improve attention, its impact on learning depend on the
classroom organization. Also see Powell, et al. (1998).57 well as started school seven months
earlier.

c
c
c c
56. De Pee, S., M. Bloem and L. Kiess, 2000: Evaluating Food-Based Programmes for their
Reduction of Vitamin A Deficiency and Its Consequences, a

, 21(2):
232-238. Deutsch, R., 1999: How Early Childhood Interventions Can Reduce Inequality: An
Overview of Recent Findings, Inter -American Development Bank, Washington, DC, mimeo.
57. Devarajan, S., L. Squire and S. Suthiwart -Narueput, 1997: Beyond Rate of Return:
Reorienting Project Appraisal,
" 12(1): 35-46.
Finally,Behrman, et al (2003)investigate the impact of community-level experimental
nutritional interventions in rural Guatemala on a number of aspects of education, using the
well-known INCAP longitudinal data set dating back to the initial intervention in 1969 -77
(when the subjects were 0-15 years of age) with the most recent information collected in
2002-3(when the subjects were 25-40 years of age) and find that being exposed to a randomly
available nutritional supplement when 6-24 months of age had significantly positive and
fairly substantial effects on the probability of attending school and of passing the first grade,
the grade attained by age 13 (through a combination of increasing the probability of ever
enrolling, reducing the age of enrolling, increasing the grade completion rate per year in
schooling, and reducing the dropout rate), completed schooling attainment, adult
achievement test scores, and adult Raven·s test scores.It is relatively straightforward to infer
from the impact of nutrition on schooling attainment to the productivity lost due to early
malnutrition using the substantial literature on wages and schooling.
There are hundreds of studies on the impact of schooling attainment on wages -- many of
which are surveyed in Psacharopoulos (1994) and Rosenzweig (1995). Wages, however, are
also directly influenced by cognitive ability, as well as by the appreciable influence of
cognitive ability on schooling achieved. Poorcognitive function as a child is associated with
poorer cognitive achievement as an adult, see Behrman, et al. (2003), Martorell (1995),
Martorell, Rivera and Kaplowitz (1989), Haas, et al. (1996), Martorell 1999) and Martorell,
Khan and Schroeder (1994). A series of studies show that reduced adult cognitive skills
(conditional on grades of schooling completed) directly affect earnings It is possible to use
these studies to estimate the magnitude of the productivity costs of poor nutrition.
Alderman, Hoddinott and Kinsey (2003) use the values for the returns to education and
age/job experience in the Zimbabwean manufacturing sector provided by Bigsten et al.
(2000) to infer the costs associated with poor nutrition in Zimbabwe. The loss of 0.7 grades of
schooling and the seven-month delay in starting school there translates into a 12% reduction
in lifetime earnings. Behrman and Rosenzweig (2004) take a more direct approach. They

c
c
c c
study a sample of adult identical twins in the United States and determinethat with controls
for genetic and other endowments shared by such twins (which would not be affected by
programs to increase birth weight),
the impact of LBW on schooling or wages is far larger than it appeared without such
controls (e.g., the impact on schooling attainment is estimated to be twice as large, with a
pound increase in birth weight increasing schooling attainment by ab out a third of a year). 58
77)$+(3)6$+"+3(+$)+

Before turning to possible opportunities related to hunger and malnutrition in
developingcountries, it is important to be clear about the framework for representing such
opportunities, including itsstrengths and limitations. 59

&2)1*))+3+" 77)$+
To identify which of the nine areas of interest for the Copenhagen Consensus present real
opportunities in the sense of having high expected benefit-to-cost ratios or high internal
rates of returns, information is required on time patterns of expected resource costs and
benefits, uncertainties for each of these time paths, and the appropriate discount rate. It is
also important to identify the relevance of opportunities in terms of basic policy motives of
distribution (e.g., alleviating poverty) and efficiency (e.g., narrowing the difference between
private and social rates of return to an action). perspective about what underlies the benefit-
to-cost ratios or internal rates of returns on which the Copenhagen Consensus Expert Group
will have to make their ranking, it is useful to review here some of the major components of
such calculations.14
5+*Estimating the present discounted value of benefits due t o an intervention is
difficult for the following reasons: (1) Most interventions have multiple impacts, some
positive and cutting across the different challenges considered here and some negative,
because they affect the basic constraints under which entities such as households and firms
and farms are making their decisions -- thus resulting in what Rosenzweig and Wolpin
(1982) call ´the unintended consequencesµ of programs and policies. (2) Most economic
evaluations of the magnitudes of impacts must be made on the basis of imperfect non-
experimental data, which raises problems

c
c
c c
58. Dunn, J., 2003: Iodine Should Be Routinely Added to Complementary Foods, "

%

, 133:3008S-3010S.
59. English R. M., J.C. Badcock, T. Giay, T. Ngu, A -M Waters, and S A Bennett, 1997: Effect of
nutritionimprovement project on morbidity from infectious diseases in preschool children in
Vietnam: comparisonwith control commune, "
, 315: 1122-1125.
of estimation due to endogeneity of right-side variables, unobserved heterogeneity, and
selected subsamples60.Though there are a number of studies that provide evidence that
controlling for such estimation problems can change impact estimates substantially (and
even, in some cases, reverse the signs!), far too often associations in micro or in aggregate
data are interpreted as if they reflect causal effects. (3) Even if the impacts are well -
measured, they are not always in monetary or easily monetized terms. 61
But they must be expressed in such terms in order to be made comparable to the costs and to
the impacts of other changes. There are various means of doing so that areused in the
literature. For example, options that are used to address the vexing problem of assigning a
value to postponing mortality include the present discounted value of the estimated
productivity of the individual that is foregone due to early mortality or the resource cost of
the most effective alternative means of postponing such mortality. We utilize the latter
approach below.16 (4). Most impact estimates do not relate well to the basic policy motives
of efficiency and distribution. In particular, the efficiencypolicy motive pertains to
differences between private and social impacts due; for example, to externalities for which
reason there may be an efficiency argument for public subsidies for an action62. But most
micro estimates present only private impacts and most aggregate estimates present only
total impacts, so neither provides a basis for judging the efficiency case for interventions. In
addition, impacts often are spaced over time, with different time patterns for different
impacts. Because there is an opportunity cost in terms of waiting, the benefits must be
discounted. But while the 1Knowles and Behrman (2003) discusses these issues at some
length. The health issues we consider are often addressed using disability adjusted life years
(DALYS).
The approach can be adapted to accommodate assumptions on discount rates and to
consider changes in productivity conceptually the same as with disabilities. While the DALY
approach is generally used for cost effectiveness comparisons, it can be converted to cost-
60. Fan, S., 2002: Agricultural research and urban poverty in India. EPTD Discussion Paper
No. 94, International Food Policy Research Institute, Washington DC.
c
c
c c
61. Fan, S., C. Fang and X. Zhang, 2001: How agricultural research affects urban poverty in
developing countries: The case of China. EPTD Discussion Paper No. 83. International Food
Policy Research Institute, Washington DC.
62.Feldstein, M., 1995: Tax Avoidance and the Deadweight Loss of the Income Tax, National
Bureau of Economic Research Working Paper No. 5055, Cambridge, MA.
benefit analysis with assumptions similar, but not identical to those used here. See, for
example, Levin et al. (1993). 63
There are basic questions about how best to evaluate DALYS that are parallel to the
questions discussed in the text for mortality. That is, what we argue is the pr eferred
alternative for evaluating a DALY due to a particular intervention is to usethe least-cost
alternative means of attaining one DALY, but in some studies (e.g., see discussion of
Opportunity 4 in Section 3.2 below), DALYS are evaluated by per capita income, which
leads to much larger estimated benefits. 15 Some of these issues also may be problems with
experimental data, depending on the details of the experiments.
There are a relatively few cases in which impact evaluations in developing countries are
based on what appear to be good experiments such as those associated with the Mexican
rural anti-poverty human resources investment program PROGRESA (e.g., Behrman,
Sengupta and Todd 2003, Behrman and Hoddinott 2002, Gertler, et al. 2003)
or smaller-scale experiments such as in NGO schools in Kenya (e.g., Miguel and Kremer
2002, 2004) or differential nutritional supplements among four rural Guatemalan
communities (e.g., Martorell 1995, Behrman, et al. 2003) or with random assignment of
vouchers among poor student applicants in urban Colombia (e.g., Angrist, et al. 2002).
This strategy requires the identification of the least cost alternative action that has exactly
the same effects as the one being valued. This is a strong requirement. But the alternatives
also require strong assumptions.64While there are examples in the literature of basing such
assessments as the value of reducing mortality on expected lifetime earnings, for example,
this methodology is fraught with pitfallsincludingthe impli cit ranking of the value of life as
a function of wages within and acrosscountries.In addition, assigning a value in proportion
to earnings does not net out consumptionfrom these earnings. 65
63. Fiedler, J., 2003: A Cost Analysis of the Honduras Community-Based Integrated Child
Care Program. World Bank, Health Nutrition and Population Discussion Paper, Washington
DC.

c
c
c c
64. Fielder, J., 2000:The Nepal National Vitamin A Program: Prototype to Emulate or Donor
65.Fiedler, J., D. Dado, H. Maglalang, N. Juban, M. Capistrano and M. Magpantay, 2000:
Cost Analysis as a Vitamin A Program Design and Evaluation Tool: A Case Study of the
Philippines,
, 51:223-242. Enclave? ,
, , 4(2):
145-156.
Mathematics of discounting are well-understood, there does not seem to be agreement on
what discount rate is appropriate.
The choice of the discount rate can make a considerable difference for investments for which
there are considerable lags (e.g., the impact of improved
nutrition on adult health
five or six decades later). Table 2 provides some illustrations of the present discounted
values (PDV) of animpact of $1000 over different time horizons and alternative discount
rates.66 The PDV of $1000 received 50 years later is $608.04 with a discount rate of 1%,
$371.53 with a discount rate of 2%, $228.11 with a discount rate of 3%, $87.20 with a discount
rate of 5%, and $8.52 with a discount rate of 10%. For impacts with lags of this duration, the
benefit (and thus the benefit -cost ratio) can be tripled by changingthe discount rate from 5%
to 3% or increased by ten by changing the discount rate from 10% to 5%.
Thus whether a project with this benefit is a great choice or a lousy choice may depend
critically on the discount rate that is used. 67 For lags that are less long, the PDVs vary less
with the discount rates, but they still may vary substantially even with lags of 5 -10 years.
Therefore any effort to compare projects across various domains (challenges) may end up
only reflecting the different discount rates that differentanalysts choose rather than the true
merits of the alternatives. )The present discounted value of the costs should equal the
marginalresources needed forthe intervention. Since a significant share of the total costs of
an intervention typically is incurred at the time of the intervention, the question of what
discount rate is appropriate is not likely to be as important for the costs as for benefits. 68
Nevertheless, discounting still may be of some importance because the initial costs may be
spread over some time and because there may be important recurring costs. Two other
pointsabout therelevant costs are important to mention because they are not always
followed in the literature. First, the relevant costs are resource costs that do not include
transfers even though transfers may be important parts of public sector budgetary costs.
66. FAO (United Nations Food and Agriculture Organization), 2003: j

% %

-001, FAO, Rome.
c
c
c c
67.Foster, A. and M. Rosenzweig, 1993a: Information, Learning, and Wage Rates in Low-
Income Rural Areas, "

%
28(4): 759-79.
68. Foster, A. and M. Rosenzweig, 1994: A test for moral hazard in the labor market: Effort,
health and calorie consumption, #
%)

, 76(2): 213-227.
Second, the relevant costs arethose experienced by society, not only by the publi c sector. The
private costs often include, for example, the time costs for members of society and the
distortionary costs of raising revenues for interventions and of making the interventions 69.
Both may be considerable. For instance, a major share of the costs of timeintensive
interventions such as schooling or training often is the opportunity cost of time not spent in
productive activities. For another example, it has been estimated that the distortionary cost
(often called the "deadweight loss") of raising a dollar of tax revenue in the United States
and in other countries rangesfrom $0.17 to $0.85, depending on the type of tax used (e.g.,
Ballard, Shoven and Whalley 1985, Feldstein 1995, van der Gaag and Tan 1997). %

Three further points, finally, are common to both benefits and costs. First, both are

. They both may depend importantly, for example, on relative prices, the
stateof the economy, and the nature of the health environment and broader aspects of the
environment. 70 High return policies in a society with high prevalence of a particular disease,
for example, may not be very effective in a society with low prevalence of that disease.
Second, for both benefits and costs there may be %% that have important
feedbacks through changing relative scarcities and market prices,particularly in economies
that are more closed to international trade. 71
Third, serious assessments of impacts and of costs must incorporate
% #
of
individuals, collective units such as households (with intrahousehold distribution possibly
affecting distribution by gender and by generation), and other entities ² including
governments (due to the endogeneity of policy choices). In a nutshell, evalua tion of
opportunities is very difficult and is not likely to lead to a set ofrecommendations that are
universally applicable in all developing country contexts.
Claims that there are such interventions are likely to be misleading and costly in terms of
their use of scarce resources72
69. Galloway, R. and J. McGuire, 1996:Women Need?

# , 54(10): 318-323.
70. Garcia-Casal, M. and M. Layrisse, 2002: Iron fortification of Flour in Venezuela,

# , 60(7): S26-29.

c
c
c c
71. Gera, T. and H.P.S. Sachdev, 2002: Effect of Iron Supplementation in Incidence of
Infectious Illness in Children: Systematic Review, "
, 4:1142-54.
72. Gertler, P., J. Rivera, S.Levy, and J.Sepulveda, 2003: Mexico·s PROGRESA: Using a
Poverty Alleviation Program as an Incentive for Poor Families to Invest in Child Health,
(forthcoming).
Our 17 Operationalizing thepoint about transfers, however, may be difficult in many
circumstances. Transfers in terms of money are fairly straightforward. In-kind transfers are
more ambiguous. Goods such as foodDaily Versus Weekly: How Many Iron Pills Do
Pregnantprovided in-kind typically are (appropriately) considered transfers, but in-kind
provision of services (e.g., subsidized schooling or health clinics) often are not considered
transfers ² though the real distinction is not clear. Discussion of opportunities is subject to
these important caveats and for Opportunity 1 ² reducing the prevalence of LBW ² we
explore how sensitive some of the estimates are to underlying assumptions toillustrate some
of these points. But it should be noted that these qualifications and problems are not unique
to this Challenge paper, but are pervasive for all the Challenge papers. 73

)$ 77)$+*)3$+"6$+"+3(+$)+
77)$+ -3$+",%(+)*5!
Many of the 12 million LBW infants born each year die at young ages, contributing
significantly to neonatal mortality, this makes up the largest proportion of infant mortality
in many developing countries. Unfortunately, rates of LBW have remained relatively static
in recent decades. Because LBW infants are 40% more likely to die in the neonatal period
than their normal weight counterparts, addressing LBW is essential to achieve reductions in
infant mortality. 74 LBW also may be important in light of new evidence that shows that LBW
infants may have an increased risk of cardiovascular disease, diabetes and hypertension
later in life. LBW may also be an intergenerational problem because LBW girlswho survive
tend to be undernourished when pregnant with relatively high incidence of LBW children. 75
3$+" +*+ &)( The probability of infant mortality is estimated to be

significantly higher for LBW than for non -LBW infants. The studies by Conley, Strully and
Bennett (2003) and Ashworth (1998) reviewed in Section 2 show an elevated risk of
infantmortality associated with LBW that translates into fairly large differences in mortality
rates given the relatively high mortality rates in many developing countries.

c
c
c c
73. Glewwe, P., 1999: j

%
2 $ $

%+, St. Martin's Press, NewYork.
74. Glewwe, P., and H. Jacoby, 1994: Student Achievement and Schooling Choice in Low-
income Countries.
75. Evidence from Ghana, "

%
, 29(3): 843-864.
The Indian and Guatemalan samples that Ashworth summarizes, for example,have neonatal
morality rates of from 21 to 39 per 1000 births and post-neonatal mortality rates per 1000
neonatal survivors of from 25.3 to 60.0.
Moreover, many of the LBW children who survive infancy suffer cognitive and neurological
impairment and are stunted as adolescentsand adults. Thus, in addition to contributing to
excess mortality, LBW is associated with lower productivity in a range of economic and
other activities.76LBW also may be important in light of new evidence that shows that LBW
infants may have an increased risk of cardiovascular disease, diabetes and hypertension
later in life. LBW may also be an intergenerational problem because LBW girls who survive
tend to be undernourished when pregnant with relatively high incidence of LBW children. 77
How to value a life saved is a big question about which there is a range of views. One
possibility, as noted above, is to use the resource costs of alternative means of saving a life.
Summers (1992) suggests that World Bank estimates of the cost of saving a life through
measlesimmunization were on the order of magnitude of $800 per life saved in the early
1990s.
3$+" 33)+())*+)+(&3( +)+3$)5! This is
the sum ofthe extra neo-natal care in hospitals as well as the additional costs of out-patient
care. The former is the product of the costs of a day of a hospital stay times the number of
additional days on average for children born weighing less than 2.5 kilograms and who are
born in hospitals. Given that the share of children born in hospitals is small in most
developing countries, under the assumption that these costs are incurred only for children
76. Glewwe, P., and H. Jacoby, 1995: An Economic Analysis of Delayed Primary School
Enrollment and Childhood Malnutrition in a Low Income Country, %
%)

, 77(1): 156-69. Glewwe, P., H. Jacoby, and E. King, 2001: Early Childhood Nutrition
and Academic Achievement: A Longitudinal Analysis, "

%,)

, 81(3): 345-
368.

c
c
c c
Enrollment and Childhood Malnutrition in a Low Income Country, &
%)

, 77(1): 156-69. Glewwe, P., H. Jacoby, and E. King, 2001: Early Childhood Nu trition
and Academic Achievement: A Longitudinal Analysis, "

%,)

, 81(3): 345-
368.
born in hospitals, the contribution of this component to the total may be small, even though
its costs are not discounted over many years.78

3$+"33)+())*$':$+((++3(3&3(
*) +*+ +3 (3+ 3$ ) 5! Ashworth (1998) reports a regular pattern of
increased morbidity with lower birth weights, particularly in the first two years of life. For
example, days with diarrhea among LBW children 0-6 months increase 33% compared to
normal birth weight in Brazil and 60% for children 0-59 months in for non-LBW infants 32.2,
so the difference is 32.2 or a probability of 0.032. Together these calculations imply thata
shift of an infant from LBW to non LBW reduces the probability of mortality in such a
population by about 0.078.

77)$+ -m&7)%+"m+*++3(3 $)++3#($%
5*3+"
,)&))+
The nutritional literature emphasizes that under nutrition is most common and severe
duringperiods of greatest vulnerability (Martorell 1997, UNICEF 1998). One such period is

.
Opportunity 1 addresses that period. A second vulnerable period is the first two years or so
of life. Youngchildren have high nutritional requirements, in part because they are growing
so fast. Unfortunately, thediets commonly offered to young children in developing countries
to complement breast milk are of lowquality (i.e., they have low energy and nutrient
density), and as a result, multiple nutrient deficiencies arecommon. Young children are also
very susceptibleto infections because their immune systems are bothdevelopmentally
immature andcompromised by poor nutrition. 79

c
c
c c
Enrollment andChildhood Malnutrition in a Low Income Country, '
%)

, 77(1): 156-69.
79. Glick. P., and D. Sahn, 1997: Gender and Education Impacts on Employment and
Earnings in WestAfrica: Evidence From Guinea, )

& '
$ $
45(4): 793-823.
In poor countries, foods and liquids areoften contaminated and are thus key sources of
frequent infections which both reduce appetite andincrease metabolic demands.
Furthermore, in many societies, suboptimal traditional remedies forchildhood infections,
including withholding of foods and breast milk, are common. Thus infection
andmalnutrition reinforce each other. The second opportunity that we emphasize is directed
towardsimproving the nutrition of infants and young children.Opportunity 2 differs from
Opportunity 1 more with regard to type of interventions that may bepromoted than in
potential benefits. To a fair degree the expected gains from improved child nutrition arethe
same as those for changes in LBW. In both cases, there are immediate benefits in terms of
reducedmortality and health care costs as well as subsequent benefits in improved
productivity from associatedcognitive development and increased stature. Even
intergeneration transmission of nutritional shocks inchildhood is perceived to be similar to
those from LBW. Of the seven categories of benefits listed forOpportunity 1, perhaps the
two that differ most from the current opportunity are numbers 2 Reduced neonatal care
associated with LBW.80
Neonatal care obviously isnot relevant for children who are older than the neonatal phase,
though childhood nutrition and feedingpatterns also are believed to influence subsequent
chronic disease. But, as noted in the discussion ofOpportunity 1, with a 5% discount rate,
these two benefits are only about a tenth of the overall benefit. 81 Infant and child nutrition is
often depicted as the outcome of the combination of appropriate foodand health inputs
mediated through child care practices. Several implications follow from thisobservation.
77)$+ -3$+",%(+)*m)+*+ +&+3m)3+

&++3;+*+
Benefits from Reducing Micronutrient Deficiencies
: The benefits from reducing iodine
deficiency are conceptually similar to those fromreducing LBW ² indeed; an important
means of reducing the LBW is to reduce iodine deficiencies inpopulations in which the
c
c
c c
80. Goh, C. 2002: Combating Iodine Deficiency: Lessons from China, Indonesia and
Madagascar, a

, (3): 280-291.
81. Golden, M.H. 1994: Is Complete Catch-Up Growth Possible for Stunted Malnourished
Children? )
"

%$
, 48: S58-S70.
prevalence of such deficiencies is high. Both for reducing LBW and reducingiodine
deficiency, the main concerns are infant mortality and irreversible impairment of mental
capacitieswhich manifest in lower productivity later in life. 82 That is, although iodine is a
required nutrientthroughout life, most documented consequences of deficiencies are from
prenatal and early child efficiencies (Dunn, 2003). This evidence comes from epidemiological
studies such as the 13.5 difference in IQ between deficient and normal individuals cited
above as well as maternal supplementation studies(reviewed in Grantham-McGregor,
Fernald and Sethuraman, 1999b and Black, 2003).
The evidence onwhether the effects of iodine deficiencies in utero can be reversed or
whether supplementation or fortification to children can improve performance is less
conclusive than the evidence on interventions before birth; though some recent evidence is
encouraging (von den Briel, West, and Bleichrodt, 2000) it isgenerally recognized that the
earlier an iodine deficiency can be addressed the greater the impact.In addition to its
influence on cognitive development iodine deficiencies have been linked toincreased child
mortality with two randomized trials of intra-muscular injections in extreme iodinedeficient
areas showing that treatment reduced infant and child mortality by 30% (Rouse,
2003).83 Similarly, a trial using iodine treated irrigation water reduced infant mortality by half
in three villages inChina (DeLong et al., 1997). In addition, deficiencies lead to lower birth
weight or subsequent weight forage; . : Prior to the 1980s vitamin A
supplementation campaigns were mainly advocated forcommunities with clinical symptoms
indicating risk of blindness. However, in light of meta -analysis offield trials of mass
supplementation to children 6 to 59 months of age that indicated an overall reductionin
child mortality by 25-35% (Beaton, et al. 1993, Fawzi, et al. 1993) vitamin
A promotion is now routinein low income countries. These and subsequent trials were less
consistent in the rates of reduction in morbidity with greater reduction in severity than
incidence of illness (Villamor and Fawzi, 2000).Impacts on morbidity are also mediated by
presence or absence of other deficiencies, for example zinc.As with morbidity, the evidence

c
c
c c
on vitamin A interventions and growth is mixed with little impact of supplementation on

growth of moderately deficient children. Children who are severely malnourished or who
also have serious infections, however, may benefit more from supplementation (Rivera, et
al., 2003).
82. Haas, J, S. Murdoch, J. Rivera, and R. Martorell. 1996: Early nutrition and later physical
work capacity.
( 54: S41-S48.
While initial evidence on prophylactic supplementation was based on provision to children
6months or older, targeting of new born children may have as significant an impact on
infant mortalitywith significant impact on infant mortality reported with supplementation
as early as the first two daysafter delivery (Rahmathullah, et al., 2003). 83
Moreover, although current recommendations for post-partum supplementation of mothers
are aimed at improving the vitamin A status of the child, one large trial supplementing
pregnant women with weekly low doses of vitamin A showed a marked decrease in
maternal mortality in Nepal (West, et al., 1999). Additional trials are underway 84.
In addition to concerns for vitamin A status of young children there are recommended
intakes of vitamin A for older children as well as adults. Indeed, supplements are regularly
provided to school-age children, for example, in school feeding programs. However, while
the cost of this supplementation maybe low at the margin, there is little evidence on the
economic returns to such programs in terms of reduced clinical costs or increased
productivity. 85
Changes in rates of vitamin A deficiency do not correlate well with changes in mortality.
Forexample, improvements in control of exophthalmia and, presumably, mortality in Nepal
and Bangladeshhave not been accompanied by reduction in vitamin A deficiencies using
serologic cutoffs (West, 2002).
Thus, the estimates of benefits of prophylactic provision of Vitamin A or its precursors used
here will bephased in terms of an illustrative benefit of reaching an at risk individual in a
program rather than in terms of a reduction in a deficiency . This approach differs
slightly from how we estimate benefits in the case of LBW (measured in terms of LBW
prevented). This affects mainly how to interpret the cost per unit of benefit. These benefits
are usually thought to be mainly from changes in mortality. While, as mentioned,
acommonly mentioned figure for reductions in child mortality is 25-30%, it is also likely that

c
c
c c
83. Hack. M. 1998: Effects of Intrauterine Growth Retardation on Mental performance and
Behavior: Outcomes During Adolescence and Adulthood, )
"

% $

, 52:S65-71.
84. Hack M., D. Flannery, M. Schluchter, L. Cartar, E. Borawski and N. Klein. 2002:
Outcomes in YoungAdulthood for Very-Low birth weight Infants, ) "

%
, 346(3): 149-157.
85. Haddad, L. and H. Bouis, 1991: The Impact of Nutritional Status on Agricultural
Productivity: Wage
the children missed in any wide-scale program are also likely to have higher than average
risk ofmortality (and also that the marginal costs of reaching these children are higher than
average costs).Hence, we use a conservative 10% reduction in expected mortality. 86
This reduction will be for infant pluschild mortality for programs reaching newborns or
their mothers and for incremental child mortality(only) for children reached after one year of
age. We do not assume any reduction in costs of outpatientor clinic care for survivors.
Fielder (2000) reports one case of blindness prevented for every seven deathsaverted in the
vitamin A prophylactic program in Nepal. This ratio may be higher than elsewhere so
weassume a one-to-ten ratio below to estimate the impact on lifetime productivity of
reducing blindness.

m)+ The estimation of the benefits of reducing iron deficiencies must incorporate the fact
thatiron deficiency anemia can affect adult worker productivity directly, as discussed in
Section 2, as well asthrough its impairment of child development. We have indicated the
gains to iron supplementation interms of LBW in the discussion of Opportunity 1.
Regardless of birth weight, iron deficiency of youngchildren appears to have irreversible
consequences on development; Grantham-McGregor, Fernald andSethuraman (1999b) find
inconclusive evidence that subsequent treatment of young children can offsetthe
consequences of earlier deprivation though prophylactic treatment may be beneficial.
However, theinconclusiveevidence from interventions is, in part, a consequence of ethical
considerations that limit long-term interventions. In a review similar to our study, Ross and
Horton (2003) find that a halfstandard deviation reduction in IQ due to iron deficiency
anemia is consistent with the overall evidence(indeed, the most detailed long-term
intervention they cite implies a one standard deviation diminution).In addition to reduced
cognitive capacity, iron deficiency can also affect schooling and hencefuture productivity by
lowering effort and attendance; Kremer and Miguel (2004) found that treatment forworms
c
c
c c
reduced primary school absenteeism by 25% in Kenya. Similarly, productivity in adults can
beaffected by reduced capacity for effort. Based on their review of published studies ² and
hence they donot include the very recent study of benefits by Thomas, et al. (2004) - Ross
and Horton assume thisproductivity loss is 5% for blue collar workers and 17% for heavy
manual labor.
86. Evidence from the Philippines, %

%)

53(1): 45-68.
;+ Evidence from meta-analyses indicate that zinc supplementation of young children
has anappreciable impact on growth with incremental increases in height or weight
averaging over 0.3 standard

) )* 3$+" m)+ *+ +& Supplementation to address iron deficiency
anemia ismore difficult than for iodine or vitamin A deficiency since it is not possible to
provide mega-doses.Thus, with the exception of providing supplements to school children
or other groups that assembleregularly, a health worker cannot easily administer the desired
treatment and must rely on patientcompliance. Small dose supplements have fewer side
effects and greater biological effectiveness whentaken regularly but this increases the cost of
treatment. However, as mentioned above, costs of regularmonitoring of compliance may be
less than the benefits when targeting to pregnant women in high densitypopulations of
South Asia where rates of both anemia and LBW are high. The evidence cited pertains
tochild survival but it is likely that similar programs will reduce maternal mortality
although prospectivestudies are limited by ethical considerations as well as sample size
requirements (Allen, 2000).
Due to problem of compliance, an intermittent schedule has been recommended, often in
terms ofweekly instead of daily supplements. Some of these studies have found little
difference in efficacy asmeasured by serological indicators (Beard, 2000, Galloway and
McGuire, 1996).
77)$+ 0-m+%&+++)()" +%()7+")$+ "$($
Definition and description of opportunity: Approximately 798 million people are
consideredhungry. Further, inadequate access to food ² both in terms of macro and
micronutrients ² is associatedwith malnutrition amongst children and adults. Technological
improvements in developing countryagriculture may have important effects on reducing
hunger and malnutrition. In this opportunity, we focuson technological change embodied in
c
c
c c
improved cultivars - improved genetic material in seeds that permitsincreased grain yield;
higher levels of micro-nutrients; greater responsiveness to other inputs such asfertilizer or
water; reduced need for complementary inputs such as pesticides and herbicides;
greatertolerance to pests, droughts or other stresses; shorter maturity times; improved post-
harvest qualities (e.g.greater resistance to disease or pests when stored) and improved taste.
A broader definition oftechnological improvements would add to these improved
agronomic practices such as integrated pestmanagement whereby pests are controlled
through crop rotation and the careful use of natural predatorsand the use of trees for shade,
green fertilizer and nitrogen fixation in soils; and improvements incomplementary inputs
such as fertilizers and pesticides (Kerr and Kolavalli, 1999).How does this opportunity
partially solve the challenge. 87
3$+"))**))3'+*+*))3)+$&+')$(+3$'+
Increases in output can lead to reductions in food prices thus benefiting net food consumers
in both ruraland urban areas. 88 Even modest price changes can generate significant welfare
gains where food purchases,especially staples, comprise a large share of household
budgets.The magnitude of the price fall willdepend on the price elasticity of demand, the
existence and extent of price controls on food, and the extentto which local food markets are
integrated into regional, national and global food markets. It will also beaffected by changes
in demandbrought about by population and income growth. Again, there isconsiderable
evidence fromAsia (Bangladesh, India, the Philippines) that food prices have
graduallydeclined as higher yielding cultivars have spread (Byerlee and Moha, 1993, Quizon
and Binswanger,1986, Warr and Coxhead, 1993).89 In China, Fan, Fang and Zhang (2001)
estimate that investment inagricultural research led to higher output and, in turn lowered
food prices that account for 30% of thereduction in urban poverty between 1992 and 1998.
87. Haddad, L., H. Alderman, S. Appleton, L. Song, and Y. Yohannes. 2003: Reducing Child
Malnutrition: How Far Does Income Growth Take Us?
)

) 17(1):
107-131.

c
c
c c
88. Harberger, A., 1997: New Frontiers in Project Evaluation? A Comment on Devarajan,
Squire andSuthiwart-Narueput, j
* 12(1): 73-79.
89. Heckman, J. J. 1997: Instrumental Variables: A Study of Implicit Behavioral Assumptions

in One WidelyUsed Estimator, "

%
32*3): 441-62.

'
! &(+$)++++"m+379
India has a higher prevalence of child malnutrition, as manifested in stuntingand
underweight, than any other large country and was home to about one-third of
allmalnourished children in the world in the early 2000s. There are, however,
substantialinter-state differences in child malnutrition and also in the (generally meager)
progressmade since the early 1990s. The persistence of widespread malnutrition may
seemsurprising considering the recent overall shining performance of the Indian economy.
Between 1993 and 2006 net state domestic product per capita nearly doubled in the wake of
4.5% average annual growth. The main objective of this paper is to identify thereasons why
rapid economic growth has failed to reduce malnutrition more substantially.The methods
used are OLS, instrument-variable, fixed-effect and first-difference regression analyses on
the basis of panel data at the level of states in India. The results suggest that the persistence
of malnutrition is mainly explained by modest poverty reduction >despite high overall
economic growth >due to minuscule factorproductivity and income growth in the
agricultural sector, still employing 54% of theIndian labour force. Widespread rural female
illiteracy and restricted autonomy forwomen are other significant explanations.
90. Índia is shiningµ was the ubiquitous slogan boasted by the incumbent National
Democratic Alliance (NDA) in its multi-billion dollar media campaign in the run-up tothe
national elections in 2004. The NDA lost the election and an often acclaimed reason is that
millions of poor and malnourished Indians felt they had been left in the sh ade.1 Smith et al.
(2003; 2005a) analyze the role of women autonomy for alleviating child malnutrition and

c
c
c c
find empirical support for the women subjugation hypothesis, especially in South Asia, but
the relative role of income poverty is not explicitly studied.

m+)3$)+
In the early 1990s, about half of the pre -school children in India were malnourished,
asmeasured by being stunted or underweight for age. At the time, several other countries in
South Asia and Sub-Saharan Africa had similar levels of child malnutrition. Theprevalence
of child stunting and underweight in India has declined since then, but at aslower pace than
in most other developing countries. In years around 2000, the latest date for which estimates
are available for sufficiently many countries to enable meaningful comparison, only a few,
much smaller countries had a higher incidence of child malnutrition than India (WHO
2007a). As late as in 2005/06, 46% of all young Indian children were underweight for age
and 38% were stunted. The high and persistent incidence of child malnutrition may seem
surprising considering that India has done remarkably well in economic terms since the
policy reform process gained momentum in the early 1990s. Between 1993 and 2006, net
state domestic product per capita (NSDP/C) grew by 4.5% per year on average, signifying
nearly a doubling of real income. Despite this shining overall economic performance, the
prevalence of child stunting and underweight dropped by 23 and 12% only over the 13years
(or by 8.7 and 5.4 percentage points).These rates of decline look modest in comparison to
China, where child stunting fell from 33 to 10% between 1992 and 2005 and child
underweight was practically eliminated. India·s progress in reducing child underweight
since the early 1990s has been only marginally better than in Sub-Saharan Africa, a region
with high and persistent child malnutrition, but economically stagnant. The key question
addressed in this paper is why high overall economic growth has failed to bring about a
more rapid alleviation of child malnutrition in India. The main explanation advanced in the
earlier literature is the subdued position of women (mothers). This hypothesis, the so-called

c
c
c c
Asian Enigma Syndrome (Ramalingaswami et al 1996), is widely adhered to, but firm
quantitative evidence of the role of female subjugation relative to that of income poverty is
scarce.91 The research tools applied are panel and first-difference regression analyses basedon
data at the level of Indian states. In addition to trying to quantify the relative impact of
thefundamental causes of child malnutrition, income poverty
91. Smith et al. (2003; 2005a) analyses the role of women autonomy for alleviating child
malnutrition and find empirical support for the women subjugation hypothesis, especially in
South Asia, but the relative role of income poverty is not explicitly studied.
and female education and autonomy, a further aim is to identify pathways through which
these variables affect children·s nutritional status. In that c ontext, we will also examine how
child and maternal malnutrition are inter-related.
Some limitation in the scope of the study should be mentioned upfront. The first is that
gender differences in child malnutrition will be beyond the focus of the analysis.
Moreover, the study is confined to investigating causes of malnutrition. The consequences
for the individual, higher burden of disease, elevated mortality risk, retarded cognitive
development and impaired labour productivity later in life >or for society in the form of loss
of human capital and slower economic growth >will not be addressed, but have recently
been analyzed elsewhere.92
The next section presents a brief overview of child nutritional status in India as a whole and
by state. The theoretical framework and the econometric models to be estimated are
presented in section 3. Variable definitions and measurements are outlined in section 4.
Results from panel OLS, instrument-variable, fixed effect and first-difference regressions are
presented in section 5. Section 6 reports results from regressions aimed at identifying the
pathways thought which child malnutrition is caused. Section 7 aims at disentangling the
extent to which the small reduction in child malnutrition is attributable to weak response to
changes in determining variables and to small changes in these variables themselves.
Qualifications and robustness tests are provided in section 8. Section 9 summarizes and
section 10 discusses policy implications.
+)%%2)*&(+$)++m+3+3'
Between the first two surveys, child stunting was almost flat, while underweight dropped
by 4 percentage points. Between 1998/99 and 2005/06 it was the other way around: child
stunting declined by 7 percentage points, while underw eight remained practically unaltered.

c
c
c c
hesedifferent developments provide an indication that child stunting and underweight may
havepartly different causesÿ a possibility to be examined.
92.Differences in the formation of child malnutrition in rural and urban areas have recently
been investigated by Smith et al (2005b) in a cross-country study.
For recent studies ofgender differences in Indiaÿ see Pande (2003) and Tarozzi and Mahajan
(2006).
India is far from being a homogenous country in terms of malnutrition. Childstunting and
underweight have persistently been more prevalent in some of the landlocked northern and
central states than in the rest of India (Figure 2). The rates at which the incidence of child
c
c
c c
stunting and underweight have changed also vary notably across the states. In six large
Indian statesÿ child underweight actually increased between the two most recent surveys.
The subsequent statistical analysis aims at explaining this dismal development.
stimates of the incidence of child malnutrition across households according to´wealth+

(possession of selected durable consumer goods) quintiles and maternaleducation in child
stunting and underweight in the household quintile with the lowest ´wealth+ score is more
than twice that in the highest quintile. The ratio of the incidence of child malnutrition in
households where mothers have no education to those with the highest is also above two.
These observations provide an indication that income poverty and female education are

c
c
c c
likely to explain part of the variance in the prevalence of child (and mother) malnutrition
across states in India (although they say nothing about relative-impacts).

c
c
c c
It is notableÿ thoughÿ that stunting and underweight is prevalent (20-25%) also inthe
households in the highest wealth quintile and with mothers who have more than 10years of

c
c
c c
education93 This suggests that for children to be brought up in a relativelywealthy and well
educated household is not sufficient for avoiding malnutrition. In turn,5 this indicates that
factors other than household income and female education arecontributing to child
malnutrition.The subsequent statistical analysis aims at identifying these other
causes.94Since there are no estimates of stunting and underweight by state with the new
norms for previous survey years, we use the old ones in order to accomplish inter-temporal
comparability. It is notable, though, that the choice of norms affects the estimated levels of
stunting and underweight, but ha ve little impact on the inter-state differences or the rates of
change over time, which are the concerns here. The work with establishing new norms also
entailed a renewed investigation of whether the standard international norms for child
genetic potential growth apply to Indian young children as well, which the study confirmed
(Bhandari et al 2002).95c

m5
All variables used in the subsequent econometric analysis are measured at the level of
states in India. There are several hoped-for advantages with units of observation at this
93 Behrman et al (2004); Alderman et al (2007); Grantham-McGregor et al (2007); Black et al

(2008), Victoria et al (2008); Huddinott et al (2008; Lancet); Lópes-Casanova et al (2005);
Deaton (2007), Walker et al (2007); Horton et al (2008).4 The modest decline in child stunting
and underweight between the years 1992/93,1998/99 and 2005/06, when National Family
and Health Surveys (NFHS) were carried out, is depicted in Figure 1.
94 Studies from India based on household observations include those cited in note 2 above;
also see Borooah (2004) and references cited therein, the World Bank (2004), Radhakrishna
and Ravi (2004) and Bharati et al (2008).
95 Other recent studies based on state data include Deaton and Drèze (2002), Besley and
Burgess· (2004) study of labour market performance in India and Deaton·s (2008)
investigation of the causes behind the distribution of adult heights in India.

c
c
c c
7 level of aggregation rather than unit-record data at the individual household level. 96
First, the state data allows the construction of balanced data panels that permits panel and
first difference analyses.
A second advantage with data at the level of states is that externalities can be captured more
accurately. Children·s nutritional status depends on several household-specific
characteristics, such as income and mothers· education, but also on unobservable variables
that reflect the broader environment in which the household lives and dwells. These could
be cultural norms that dictate women·s role in society, or the general level of poverty in the
area that may have bearings on the incidence of transmittable diseases and the quality of
health-care services. 97
A third advantage with aggregated data (i.e. states) is that noise of various typestends to
average out, hopefully leading to more efficient estimates. Individual household data
usually contain large random measurement errors, which induce an attenuation
bias(Wooldrige 2006).
Moreover, data for individual households also pick up short-term fluctuations in variables
that do not reflect more permanent conditions. 98
There are, however, some potential drawbacks with using data at the level of states.
One is that the limited number of observations (at the most 48 in some of the panel
regressions)reduces the degrees of freedom and hence the size of the models that can be
tested.
In the case of India, this implied that the share of stunted children aged 0 -3 years in 2005/06
jumped from 38.4 to 45.2%, while the share of underweight dropped from 45.9 to 41.8%.
The Indian states may also be too large and internally diverse. Ideally, smaller areas would
be preferred, e.g. the 593 districts (in the 2001 Census) that comprise the next layer of
administrative unit in India.
96 In the DHSs, data on some observable community characteristics are usually collected in
rural areas, but so far not in urban settings.
97 The World Bank (2004) provides charts with the incidence of child underweight at a
lower level than states in India, but acknowledges that these are not representative.
98 In 2006, the WHO changed the norms for ńormalµ height and weight for specific age and
sex.

c
c
c c
However, representative data on child stunting and underweight, as well as on most

explanatory variables, are not available at this lower level of aggregation.
The data on child nutrition status are from the three National Family Health
Surveys (NFHS 1-3) carried out in 1992/93, 1998/99 and 2005/06. Most other data are
from the large 50th, 55th and 61st National Sample Surveys (NSS) conducted in
1993/94,1999/00 and 2004/05, but also various other official sources including the Reserve
Bank of India (RBI). The bulk of the Indian state data are obtained in overlapping years
(1993, 1999 and 2005), which enable us to construct balanced data panels. A statistical
supplement to this paper gives details of the data used, more complete references and
discussions of data shortcomings (work in progress).
c
7+3+%'(
Child malnutrition will be measured alternately by the prevalence of stunting and
underweight among 0-3 year old children. Child stunting (underweight) is defined as a
height (weight) for age below 2 standard deviations from the median height (weight) of
the norm children. The age-specific estimates of stunting and underweight are derived on
the basis of the WHO/NCHS norms.8 Stunting (retarded skeletal growth) isconventionally
regarded as the most sensitive marker of long-term deprivation of micronutrients and
frequent and prolonged illness. Underweight (low mass of fat and muscletissue) reflects
calorie deficiency and more acute illness (Waterlow 1992; Shrimpton et al 2001; Lancet 2008).
Wasting (low weight for height) was also tried in a first round of regressions, but few
significant results were obtained as there is very little variation in this measu re across states
and over time.
The focus on children (and their mothers later on) is dictated by the seldom contested
perception that they are the nutritionally most vulnerable groups. This is also reflected in the
fact that anthropometric data for other population segments are not frequently collected.
Estimates of child stunting and underweight are available for all Indian states, but most of
the analysis here has to be confined to the 16 largest ones since data for many explanatory
variables are lacking for the 17 mini-states and UnionTerritories (jointly with 4% of the
Indian population).

c
c
c c
$+3&+(#7(+) %'(

,)% .
At the level of states in India, the share of the population that can afford anadequate diet is
determined mainly by average household per capita income, the interhousehold distribution
of incomes and relative (food) prices >the three main building blocks behind the poverty
estimates. Food comprises about half of total consumption expenditures of the average
Indian household and nearly three-fourths in the lowest income quartile (Sen and Himanshu
2004). Low income also constrains households· ability to feed children food with a high and
balanced micro-nutrient content as such food items, i.e. animal products, fruits and
vegetables, are invariablymore expensive than staple grains. Poverty further reduces
households· ability to demand for qualified child and m aternal health care. India is special
in the sense that three-quarters of all health expenditures are private, out of the pocket. This
proportion is higher than in almost all the other 192 countries for which the WHO (2007b)
provides estimates. Income poverty also constrains household demand for adequate
housing, sanitation and water supply.

&(((

Mother illiteracy will be used as the measure of maternalability to care well for children.
There are at least four reasons for expecting maternalilliteracy to impair the nutritional
status of their children. One is that illiterate mothers arein a disadvantaged position to
acquire and apply knowledge about appropriate health -careand feeding practices. A second
is that uneducated women are likely to be less able tocare well for themselves in terms of
nutrition and health and therefore less apt to care fortheir own children. A third is that
uneducated women marry earlier and have higherfertility (Abadian 1996; Smith et al 2003;
Smith et al 2005a). A fourth reason is, as wewill see, that illiterate women abstain from
exercising their right to vote in state election
more often than their literate peers and this affects the public provision of child health
care.
)<$)+)& .
In India, as in most countries, mothers are the chief caretakers of children, feeding them and
seeking health care when they are sick. Mother·s possibilities of undertaking these
responsibilities >given their ability >can be constrained by gender-biased cultural values.
The less autonomy or clout women have within the household and in society, the less likely

c
c
c c
it is that their own and their children·s wellbeing is prioritised in the intra-household
allocation of resources (Abadian 1996; Smith et al 2003). Ancient cultural norms that
subjugate women in India and other south Asian countries have been advanced as the main
reason why malnutrition in this region is much more prevalent than in poorer Sub -Saharan
Africa (Ramalingaswami et al 1996).99This study will investigate whether differences in
gender-related cultural values withinIndia contribute to explaining >with due controls for
other influences >the inter-statevariation in child malnutrition.
The male/female (M/F) population ratio will be used as the proxy variable for female
autonomy. This measure of ´missing womenµ is mainly determined by differences in sex-
specific death rates for different age cohorts and the sex ratio at birth (Sen 1992). A higher
than normal M/F ratio in a population reflects excess death rates of females in general in the
wake of discriminatory treatment in health care and nutrition within households, but also
sex-selective abortions (Jha et al 2006).

,6!.
In the earlier discussion of the motivations for selecting the fundamental explanatory
variables, several pathways through which these variables are hypothesised to affect child
malnutrition were sketched. Most of these pathways were examined in preparatory work
and results from two of these exercises will be presented here.
m+="+)+(77$)+)*&(+$)+
6 7): A commonly adhered to hypothesis is that when mothers are malnourished,
there is an elevated risk that their children will become malnourished as well
>intergenerational perpetuation of malnutrition (Ramalingaswami et al 1996). There are at
least three plausible reasons for expecting this. One is that underweight in pregnant women
increases the likelihood of low birth weight (LBW<2.5 kg), which in turn is a strong
predictor of underweight in infancy and early chil dhood (ACC/SCN 2000; Osmani and Sen
2003; Lancet 2008). The second reason is that malnutrition in lactating mothers reduces the
99 The prevalence of poverty in the smallest states and the UTs are simply assumed to be the
same as in neighbouring larger states (GOI 2007). In 2001, three of the largest Indian states
(Bihar, Madhya Pradesh and Uttar Pradesh) were split up into two separate states. In order
to accomplish comparability with earlier state division, all data from 2005/06 for the ńewµ
states have been merged (population weighted average) into the three óldµ states.
c
c
c c
micro-nutrient content in their breast milk, affecting infant growth adversely (Allen 2005).
The third reason is that malnourished mothers are presumably weaker and more sickly and
hence less able to care well for their off-springs (Lancet 2008).
The first link in the pathway to be tested is the determinants of malnutrit ion in mothers, as
measured by a body mass index (BMI) below 18.5. Maternal BMI failure is hypothesised to
be determined by the same fundamental variables as child stunting and underweight, i.e.
poverty, female illiteracy and autonomy. The second link that we intended to test is that
between mother underweight and LBW. Unfortunately, there are no representative
estimates of the prevalence of LBW in the Indian states (Mistra 2002).
The 30% estimate for all-India routinely provided by UNICEF and other interna tional
organisations builds on births in selected medical institutions (ACC/SCN 2000). We can
hence not test this link. The next link tested is therefore the one between stunting and 15
underweight among children and mothers with a BMI<18.5, with controls for child health
care provision, size of households, feeding practices and sanitation facilities.
The case for including health-care provision rests on findings in the epidemiological
literature; mounting evidence shows frequent and prolonged untreated illness to be one of
the most important factors behind child malnutrition (Black et al 2008; Victoria et al 2008).
Moreover, recent epidemiological research finds that micronutrient deficiency aggravates
infectious disease, which in turn leads to mal-absorption of several micro-nutrients, stifling
child growth in a vicious circle (Bhutta 2006).
The total fertility rate (TFR) i s included in the regressions as a control on the assumption that
many children in households mean that mothers have less resources andtime to care for
each child. High fertility also implies shorter birth spacing and time for mothers to
recuperate (Dewey et al 2007). Moreover, high fertility goes hand in hand with mothers
being very young and inexperienced when giving the first birth (NFHS-3 2007).10 The
feeding practice variable is included as there is almost universal agreement among experts
that from the age of six months infants should be fed supplementary solid food in order to
ensure full genetic potential skeletal growth. Adequate sanitation is important mainly for
reducing water-borne diseases such as diarrhea >a contributing cause of malnutrition (and
also the second largest killer of post-natal infants and young children in poor countries).

+3%'(&$: The two latest NFHSs provide comparable estimates of the
share of mothers with a BMI<18.5, while no such estimates were obtained in the first NFHS
c
c
c c
(1992/93). Neither were comparable data for feeding practices collected in this survey. We
hence have to restrict the panel for the 16 states to two points in time. Average household
size is measured by the total fertility rate and feeding practices by the share of 6-9 month
olds who in addition to breast milk are regularly fed solid or semi-solid food. Sanitation is
proxied by the availability of a flush toilet or a covered latrine in the home. Data on all these
variables are from the NFHSs. 100
There is no summary statistic on the share of children that is provided with qualified health
care in the Indian states. Therefore, a child-health-care index (CHCI) was constructed that
includes both preventive and curative health-care services. The index is defined as the
average of the shares of (1) births in medical facilities, (2) births assisted by health
professionals, (3) children being fully vaccinate d and (4) children brought to a health facility
when suffering diarrhea.101
The four variables are internally highly correlated, which indicates that in unison, they
should measure well the general reach and coverage of qualified health-care provision for
children. There is large variation across the Indian states in child health care provision, as
measured by this index, reflecting differences in demand and supply.In OLS panel
regressions for child stunting on the confounding variables, mother·s BMI status is
insignificant in combinations with other variables (Table 3). 102
Stunting is the most strongly associated with the child health care index. The feeding
practice variable is highly significant in the r egression without a time dummy, but
insignificant when this dummy is entered, and is hence not robust (Table 3).
100 Data from the 2005/06 NFHS show the incidence of underweight to be positively
correlated to the (higher) birth order of the children (NFHS-3, table10.1). There is most
probably causation in both directions between child malnutrition and the fertility rate, as
child stunting/underweight raise the infant mortality rate and hence fertility through the
replacement effect.
101 The definition of safe water has unfortunately changed drastically between the three
NFHSs, which means that this variable cannot be included in the regressions.
102 It is widely agreed in the epidemiological literature that stunting in infancy and during
the first two or three years of life is the main determinant of short height later in life and that
the potential for ćatch-up ´growth later is small (Shrimpton et al 2001).
c
c
c c
The time dummy itself is significant with a negative sign, indicat ing that stunting declined
over and above what is éxplainedµ by the confounding variables. In sharp contrast, in the
regressions for child underweight, mother·s nutritional status (BMI) is highly significant.
Also the TFR is highly significant with the expected positive sign, while child healthcare,
feeding practices and sanitation turn out insignificant. In this regression, the timedummy is
also insignificant.
That underweight is not associated with inadequate health care could be consistent with the
observation that weight loss (fat and lean tissue) related to disease can be reversed once the
child recuperates. That child underweight is strongly associated with (or caused by)
maternal underweight is consistent with the well-established link from underweight in
pregnant women to low birth weight and subsequent underweight in infants and young
children (ACC/SCN 2000).
That the TFR is a significantly associated with child underweight is probably because in
households with many children, mothers have less time for feeding and caring well for each
of them.That child stunting is strongly associated with lack of qualified preventive and
curative health care is consistent with the epidemiological findings t hat frequent and
prolonged untreated illness impairs infant and child skeletal growth, which is practically
irreversible.
The hypothesis ventured in section 2, that child stunting and underweight have partly
different determinants, is hence in agreement with the data.

&(%)+"+37$'((3(7)%)+
Given the apparent importance of qualified child health care for alleviating malnutrition,
especially stunting, we shall make an attempt in this section to examine in some moredetail
a presumed pathway from female literacy to child health care.

6 7) . Households· demand for qualified child health care is assumed to be
determined by income poverty, female education and autonomy. The supply of public
health-care infrastructure is assumed to be determined by government expenditures
allocated to the health sector in the states. It is further assumed that women generally give
higher priority to health care for children than their husbands (Abadian 1996; Smith et al
2003).13 Moreover, mothers are expected to be able to spend relatively more of household

c
c
c c
resources on qualified health care for their children if they are literate and have autonomy.103
There are hence four links in the pathway to be tested. The first is what determines the share
of women who exercise their right to vote in state elections relative to men. The second link
is that between state governments· allocation of resources to the health-care sector and
women/men turnout in the elections. The third link is that between provision of child health
care and state health expenditures (rupees per capita and year). 104
The final link is that between child health care and child stunting and underweight, which
was already reported in Table 3 (N=32), but a further test on a larger data set (N=48) and
partly different confounding variables will be reported from below.

+3%'(&$: The data on voting turnout refer to the latest state election
preceding the years in which the national nutritional surveys were carried out (1992/93,
1998/99 and 2005/06) as reported by the Election Commission of India. The state
government health expenditures per capita (SGHE/C) in respective year will be measured
in real 1993rupees. The data on Net State Domestic Product per capita in real terms are from
the Reserve Bank of India. The provision of qualified health care for children is measured by
the CHCI (see above).
$(: The results from OLS panel regressions of the four links in the pathway are
reported in Table 4. The panel covers the 16 largest states in India in three years. In the first
regression (column 1), the female/male turnout ratio is the dependent variable with female
illiteracy as the explanatory variable of main interest; total turnout, poverty, the M/F ratio
and time dummies are entered as controls. Female illiteracy comes out highly significant
along with total turnout and poverty, while the M/F ratio is insignificant.
103.It is notable that most micro-credit schemes in poor countries target women as the prime
lenders.
104.Finally, we assume that by voting in state elections, mothers may be able toaffect state
governments· expenditures for health care in general and for children in particular.

c
c
c c
In the second regression (column 2), state government real health expenditures per capita
(SGHE/C) is the dependent variable, regressed on F/M turnout ratio in the elections and
controls and time dummies. The F/M turnout i s highly significant, while the total turnout is
significant with a negative sign! In this regression, the NSDP/C comes out insignificant,
although with the expected positive sign. In t he third regression (column 3), with the CHCI
as the dependent variable, the SGHE/C vari able turns out significant with the expected sign.
The control variable the total fertility rate is highly significant, but the sanitation variable is
insignificant.
In columns 4 and 5, child stunting and underweight (alternately) are regressed on the CHCI
with controls and time dummies. The CHCI turns out significant in both regressions, but
more strongly so for stunting than for u nderweight.15 In the regression for stunting, the TFR
also comes out significant while not the sanitation variable. In the regressions for child
underweight it is the other way around. Sanitation is highly significant, while the TFR is
insignificant. These results indicate that the pathways through which child stunting and
underweight are caused do differ.
That female illiteracy, but not the proxy for women·s autonomy, determines the
women/male turnout in state elections is perhaps surprising from an Asian-enigma
perspective. That the voting turnout of women relative to men >but not the total turnout
>has a positive effect SGHE/C underscores, however, that empowerment of women is
important for child health. The result that SGHE/c seems to have an impact on the actual
provision of health care for children was not entirely expected. In India as a whole, the share
of health in total state expenditures is lower (3.9%) than in all but a handful of other
countries (WHO 2007b) and about three-quarters of all health expenditures are private, out
of the pocket. Considering also that state health expenditures fluctuate wildly from year to
year and actually seems to be residual item in the state budgets, the link to child health
care provision is somewhat surprising.16 Finally, the results in Table 4 are consistent with
the epidemiological findings that qualified health care is important for reducing illness and
hence for avoiding impaired skeletal growth (stunting) in infants and young children.

(7)% 3$)+-((&(+$)+3$)+
The results from the OLS and instrument-variable panel regressions (Table 1) show poverty
and female illiteracy to be highly significant det erminants of child malnutrition as
manifested in stunting and underweight. Also the first -difference regressions (Table 3) show
c
c
c c
decline in poverty and female illiteracy to reduce child stunting significantly, but not so for
underweight (which we at least partly explain by the minuscule change in underweight over
the period). That income poverty and female education matter for child nutritional status in
India is fully in line with results ob tained in most related studies, based on cross-country or
unit-record observations in individual countries, which is reassuring.
There is, however, compelling reason to go a few steps further in the strive to understand
why child stunting and, especially, underweight in I ndia have not declined more rapidly
since the early 1990s. Considering an impressive overall growth of net state domestic
product per capita in India of about 4.5% per annum, reductions of stunting and
underweight by 12 and 5 percentage points over this pe riod seem small.105
15 It should be noted the CHCI variable came out insignificant in the regression for child
underweight reported in Table 3, based on a shorter panel, but including maternal
underweight (highly significant) for which no data exist in 19 92/93.
16 The residual property and the high year-to-year fluctuations in public health
expenditures in the Indian states have been highlighted in a recent official report (NCMH
2005, p.71).
The estimated coefficients for both stunting and underweight with respect to poverty and
illiteracy in the OLS panel regressions all in the range 0.31 to 0.39 and twice as large in the IV
regressions (Table 1). In the first-difference regressions for stunting, a 1% point reduction in
the incidence of poverty translates into a decline in the prevalence of child stunting by about
0.5 percentage point (Table 4). Whether the size of these coefficients should be deemed small
or large is difficult to say since no previous comparable study of child malnutrition has used
poverty as the explanatory income variable. In the methodologically akin cross-country
investigations, the income measure used is percapita GDP, and an income-malnutrition
elasticity of around -0.50 is a standard result.17What we can say with considerable
confidence is that the growth of NSDP per capita in all-India by 4.5% per annum between
1993/94 and 2004/05 (RBI 2007) was followed by relatively little poverty reduction.
105 The supply of public health care facilities is largely determined by the budget allo cations
of state governments, which are responsible for health and education in the states, while the
central government is a minor provider of funds.

c
c
c c
Between the same years, the officially estimated incidence of poverty dropped by 8.5
percentage points only, or by 23.6% in relative terms. In relation to an accumulated increase
in the NSDP/C by 61.5%, this suggests a rough aggregate poverty-income elasticity of the
order -0.38. In a study based on data from 1958 to 1991, Datt and Ravallion (1996; 2002)
found the poverty-income elasticity in India to be between -0.75 and -1.09 depending on the
assumptions made.
Although a more detailed comparison between the pre- and post-1991 period is called for,
the reduction of poverty in response to economic growth in the Indian economy seems to
have slowed down considerably.18
Moreover, there is no correlation whatsoever between poverty reduction and cumulative
growth of net state domestic product per capita (NSDP/C) across the Indian21 states over
the period 1993/94 to 2004/05 (Figure 6). This is so irrespective of whetherthe sample is
restricted to the 16 large states included in previous regressions, or all states for which data
are available (25). All 25 states experienced some growth of NSDP/C and some poverty
reduction, but the expected negative significant correlation fails to materialise (the
regression line is positive, but insignificant ). This is in sharp contrast to cross-state
observations for India, which reveal a very strong correlation between levels of poverty and
NSDP/C. 106 This observation provides a further indication that the link between poverty
reduction and overall economic growth in India has changed since the early 1990s.107
106 Smith and Haddad (2002); Haddad et al (2003); Svedberg (2004).

107 The recent trepid poverty decline in response to accelerating growth in all-India also
seems meager in international comparison. Available estimates of poverty-income
elasticities, based on cross-country panel data, are in the -0.67 to -1.94 range, depending on
estimation method and type of data used (Kraay 2006; Loayza and Raddatz 2006). Also see
Deaton and Drèze (2002)

c
c
c c
Although a full-fledged analysis of the reasons for the meager decline in poverty in all-India
(and the lack of correlation between income growth and poverty reduction across the Indian
states) is beyond the scope of the present paperÿ one can identify four plausible contributing
factors. One is that the share of consumption expendituresÿ as broadly measured in the
national accountsÿ fell from 66% to 56% between 1990 and 2006 (ADB 2007). A second reason
is that consumption expenditures as measured in the NSS household surveys >the basis for
the poverty estimates >are only some 60-70% of consumption as measured in the national
accountsÿ and seems to have grown more slowly (Datt and Ravallion 2002).
A third reason is that the distribution of NSS household consumption expenditures has
become more uneven over time: the Gini coefficient for household expenditures increased
by 3.5 points between 1993/94 and 2004/05. This is mainly because annual growth of real
per-capita consumption expenditures in the lowest income quintile was a meager 0.85%
between the same yearsÿ while above 2% in the highest quintile (ADB 2007).
The slow growth of consumption expenditures among the poorestÿ in turnÿ is at least partly
an outcome of the fact that income growth in India has been very uneven across sectorsÿ
with agriculture as the lagging sector. During the 1993-2004 periodÿ annual growth of output
per worker in the agricultural sector averaged 0.5% while in the industry and service sectorsÿ
growth was 0.9% and 2.1%ÿ respectively (Bosworth and Collins 2008).19 The fact that more
c
c
c c
than half of the Indian labour force is still employed in the relatively stagnant agricultural
sector is hence consistent with little reduction of poverty >and of child malnutrition.20
The growth of household income has been especially low in the most populous states with
the initially highest levels of malnutrition, Bihar, Madhya Pradesh and Uttar Pradesh. This is
another contributing factor behind the tardy reduction of child malnutrition as measured by
underweight in all-India. These three states were home to more than half the underweight
children in India in 1993. As can be seen from Figure 6, the three states had very little
accumulated growth of NSDP/C and minuscule poverty reduction over the next 13 years
>and little reduction of child underweight. The slow decline of child underweight in all-
India is hence partly explained by the unfavorable overall economic performance and
persistent poverty in the most populous states. In the cross-state regressions, in which all
states carry the same weight, this size-of-state issue is hidden.
We also examined how the declines in female illiteracy and in the male/female population
ratio relate to economic growth across the Indian states. In short, the reduction of female
illiteracy between the early 1990s and mid 2000s was found to be significantly related to
growth of NSDP/C, albeit at a rather low level of significance (0.05). The minuscule change
in the M/F population ratio over t he same period was unrelated to economic growth.

>$(*)++3)'$+

In previous sections, controls for endogeneity, heterskedasticity and robustness were
made. There are a few other measurement and methodological issues to be addressed.

$()(+
The main reason why a distinction has been made between fundamental and confounding
variables is that they are expected to be highly correlated. Including confounding and
fundamental variables in the same regressions for child stunting or underweight could
hence cause problems with multicolinearity.108
108In recent years it has become increasingly clear that the sector composition of growth
matters considerably for poverty alleviation and that slow productivity growth in labour-
intensive sectors in general and in agriculture in particular, is the major reason for the failure
to reduce poverty more forcefully in the developing countries in general (Loayza and
Raddatz 2006).

c
c
c c
To test this hypothesis, the confounding variables were regressed on the fundamental
variables. The results are presented in Table5. 109
As expected, the confounding variables are significantly cor related with many of the
fundamental variables, but partly different one s. In panel OLS regressions for maternal BMI
failure, poverty and female illiteracy turn out highly significant with the expected signs. The
M/F proxy is insignificant, however, indica ting that women autonomy (as measured here) is
not affecting their nutritional status. 110
The CHCI is significantly and negatively associated with poverty (reflecting affordability)
and mother·s literacy and autonomy (reflecting maternal knowledge and priorities within
households).
The total fertility rate is strongly correlated to female illiteracy, while unrelated to poverty
and the M/F ratio, a perhaps surprising result. In the regressions for feeding practices, the
M/F21 Sanitation facility (flush toilet) is highly correlated to female illiteracy with the
expected sign, but not to any of the other fundamental variables. Finally, it is notable that all
five confounding variables are significantlycorrelated to female illiteracy >with the expected
signs.
))&+"
India probably has better, more disaggregated and more comprehensive data on most
variables that have been used in this study than a ny other low- and middle-income country.
Still, there are shortcomings. One is that in the first NFHS (1992/93), estimates of the
prevalence of stunting are missing for five of the 16 large states included in our regressions.
With an aim to have a balanced data panel and maximum degrees of 24 freedom, as well as
not ´wastingµ data on other variables, we have interpolated the 5 missing data points on
stunting, using a regression technique borrowed from the WHO (de Onis et al 200?).
Themethod relies on estimating the correlation between stunting and underweight on the
basis of the 11 states with complete data on both variables in 1992/93.
109 In the 2000 NSS labour survey 58% of the labour force was in agriculture and 54% in the
2005 survey. In absolute numbers, the agricultural labour force in India increased by 5
million, from 202 to 207 million between these years (ADB 2007).
110 A strong association between the TFR and maternal education has been found in several
studies, although an unambiguous line of causation is difficult to establish (Doepke 2004).

c
c
c c
The ensuing regression equation (highly significant) is then used to estimate stunting from
the available data on underweight in the five states.22 As a check of the reliability this
method, we re-ran all regressions for stunting omitting the five missing observations (five
states in 1992/93). The ensuing results, based on 43 observations, did not differ in any
significant manner from those obtained earlier.
A second data shortcoming is that some of the variables used in the panel regressions may
not be strictly comparable over time. A large number of scholars have pointed out
measurement anomalies in estimated household consumption expenditures in the 1999/00
NSS survey. These anomalies may have compromised the comparability of the poverty
estimates for this year and the other two survey years. This potential incomparability, in
turn, may have distorted the panel regressions reported in Tables 1 and 2. Several attempts
have been made, though, to correct the official poverty estimates from 1999/00 so as to make
them more comparable with the estimates from 1993/94 (and 2005/06). 23 Deaton·s (2003)
scrutiny of the base data and his re-estimation of poverty in 1999/00 is perhaps the most
trustworthy such undertaking. To control for the bias i n the official poverty estimates, re-ran
all previous cross-state regressions based on Deaton·s alternative poverty estimates for
1999/00. Quite reassuringly, not a single result was turned around. In fact, the differences
were negligible and the regressions based onDeaton·s poverty estimates actually turned out
marginally stronger than when based on the official ones. This could be interpreted as a
vindication of the frequent claim that the official ones do underestimate poverty in 1999/00.
There has been no questioning of the inter-temporal comparability of the poverty estimates
from 1993/94 and 2004/05 when estimates from 1999/00 are far too low and that the decline
in poverty over the 1990s to be much smaller than the official estimates suggest. Datt and
Ravallion (2002) find the official estimates to be only slightly 25 based on a uniform recall
period, which we used in the first -difference regression (Table 2).
A third potential data problem is that for some variables, alternative estimates are available.
There are at least three independent statistical sources providing estimates of female
literacy: the NFHSs, the NSS and the Censuses. If these provide diverging estimates, the use
of one particular data set could be ambiguous and induce non-robust results. When
choosing data on female (ill)literacy to be used in the present study, estimates from the three
(independent) sources were compared, for India as a whole, and for the individual states.
The comparisons revealed a very close correspondence both in terms of levels and in
c
c
c c
changes over time, indicating that no major measurement bias is likely to be found in the
(ill)literacy estimates. The decision to use the estimates from three NSSs was dictated mainly
by the fact that these surveys were conducted in the same years as the NFHSs and they
cover the same age group over time (females 7+ years).
These three NSS surveys are quintile, implying that they are especially large (thick).24
A further potential problem is that data unavailability has led to the omission of some
variables that theoretical considerations suggest should be included in the analysis.
The most obvious is the prevalence of LBW, for which no representative estimates are
available for the Indian states. The sanitation variable ´safe waterµ was also omitted because
the definition has changed drastically over the NFHSs. Relative food prices have not been
entered as an independent explanatory variable in the child malnutrition regressions. Food
price differences, across states and urban/rural areas, as well as over time, have been taken
into account indirectly since the state-, urban/rural- and times specific poverty lines used for
the estimation of poverty incorporate such differences.
c
5($+7)# %'(
Some of the proxy variables applied in this study could be poor measures of what they
intend to capture. The M/F population ratio, for instance, may not be the most adequate
measure of women·s autonomy in households and society. One problem is that the M/F
ratio could have been compromised by gender-specific emigration and/or inter-
statemigration. Kerala, for instance, is the only state where the M/F ratio is below unity
(around 0.96). This is at least partly a consequence of the fact that large numbers of men in
Kerala migrate to the Gulf states. It is notable, though, that in other Indian states, emigration
and non-seasonal inter-state migration are minuscule and has been so over the period of
concern (Lucas 1998; Srivastava and Sasikumar 2003).
An alternative proxy for female autonomy used in related literature is female labour market
participation. This variable may be poor proxy for autonomy since it can have an
independent impact on child malnutrition. It was nevertheless tried as an alternative to
M/F, but no significant results emerged. A contributing reason may be that the female
participation rate is not well defined and poorly measured in the Indian states.
Official labour market statistics set the rate at about 25% and with little variation over states
or time. It may also be that our proxy for feeding practices is too blunt, but it is the only one
available in all three NFHS.25
c
c
c c
7*)+'
In the models tested, all explanatory variables were entered linearly and independently in
the regressions. Linearity means that the marginal effect of a change in an explanatory
variable is assumed to be constant, both across different levels of the variables and across
different states. To check the validity of this assu mption, plot inspections of all regressions
were made. Only in some regressions in which the total fertility rate is included, a non-linear
specification seemed justified. Re-estimations showed, however, that the improvements in
fit were negligible. 111
The independency assumption underlying the regression model applied was tested by
introducing the interaction variable poverty*female illiteracy in the regressions. The
regressions in which poverty and female illiteracy were entered jointly, but
separately,provided better fits as measured by R-square. However, in the regressions where
the inter-action variable replaced these two variables, it was significant at a very high level.
It hence seems that whether poverty and female illiteracy are entered separately or as an
interaction variable matter little for the results.
The main objective of this study has been to explain why the reduction in child malnutrition
has been relatively small despite the impressive overall performance of the Indian economy
since the early 1990s. Although the results are only indicative, we have found that while
poverty reduction has a significant impact on the alleviation of child malnutrition in India
However, poverty decline has been modest despite high aggregate growth in the economy.
This, in turn, is at least partly a consequence of slow growth ofhousehold real consumption
expenditures among the poorest quintiles that are predominately employed in the
agricultural sector. In this sector, factor (labour) productivity growth has been much slower
111 Similar methods are frequently used by the international organisations to fill in missing
data points, e.g. by the World Bank when estimating GDP per capita in countries with
incomplete or obsolete data. The Bank·s International Comparison Program includes
benchmark estimates for less than half the about 200 countries contained in World
Development Indictors.
c
c
c c
than in the rest of the Indian economy and even declined since the late 1990s (Lal 2008).112
Female illiteracy was found to be a strong determinant of child malnutrition, which is in line
with results in earlier related literature. In all-India, female illiteracy has declined notably
since the early 1990s, from 55% to 39% in 2005. Masked behind these averages, however, is
the fact that female illiteracy fell less in the rural areas of the most populous states, with the
initially highest prevalence of child malnutrition. In these states, Bihar, Madhya Pradesh,
Rajasthan and Uttar Pradesh, rural female illiteracy was still well above 50% in 2005 and the
rural population accounts for 75-87% of the total in these states.26 Overall improvements in
female literacy has helped bring down child malnutrition according to the results reported
here, but in rural India, female illiteracy is still more than twice as high as in urban settings
(46 vs. 20% in 2005).113
The third fundamental explanatory variable for child malnutrition in this study, the M/F
population ratio, used as a proxy for women·s autonomy, was found to have a significant
impact on child stunting in the panel and first-difference regressions, but not
onunderweight. 114 In all-India, the M/F ratio has changed only marginally over time, from
107.9 in the 1991 census to 107.1 in the 2001 census, and is estimated at 106.8 in 2006 (GOI
2006b). A recent nationally representative estimate of the M/F ratio at birth found it to be
111.2, reflecting mainly gender selective abortions (Jha et al 2006). This signals that the
gender bias in India is not about to erode in the near future. 115
112 There is no agreement on the size of the bias. Sen and Himanshu (2004) claim that the
official poverty off the mark, while Deaton·s (2003) find the official estimates to be some 3-4
percentage points too low.(For additional contributions to this debate, see Deaton and Kozel
2005, Pogli et al 2005; Reddy 2007.)
113 In the NFHSs, the age group covered has changed over time /Check again/
114 In the most recent NFHS (from 2005/06), several measures of child feeding practices are
provided, such as newborns breastfed within one hour of birth, children aged 0-5 months
exclusively breastfed, and children aged 6-9 months receiving supplementary solid food. In
the earlier NFHSs, most of these data were not collected in a comparable manner.
115 In the 2001 census, the three (pre-secession) states had a joint population of 374 million,
or 36.4% of the total in India. The population in Uttar Pradesh, at 175 million, exceeded by
far the entire population in Pakistan in 2001 (141 million) and in Bangladesh (133 million).
c
c
c c
In addition to estimating the relative strength of fundamental variables behind child

malnutrition, we have examined two pathways through which children·s nutritional status
are assumed to be affected. The first is the link from mother to child nutritional status. Child
underweight (but not stunting) was found to be highly correlated to underweight among
mothers. This is in line with the world-wide observation that malnutrition in expecting
mothers is a strong predictor of LBW and subsequent underweight in infants and young
children (ACC/SCN 2000). Unfortunately, we were not able to test the LBW link directly
due to the unavailability of data on birth weights.
Underweight in mothers (BMI<18.5) themselves was found to be significantly correlated to

poverty and own illiteracy (but not the M/F ratio), the same fundamental variables behind
child underweight.
The second pathway focused on the link from female illiteracy, women·s turnout in state
election and state government expenditures on health care. The variable women/men
turnout ratio was found to be strongly associated with women literacy, but also with
poverty and total turnout (female and male). A high women/men turnout r atio was
identified as a highly significant determinant of state health expenditures. In this regression,
the total turnout was significant, but carried a negative sign. The other control variable,
NSDP/C, turned out insignificant. In the regression aimed at finding out whether qualified
child health care, as proxied by the CHCI, depends on SGHE/C, this was confirmed. Finally,
it was found that child stunting (and less so underweight) is strongly associated with the
provision of health care as measured by the CHCI.
In the regressions for confounding variables on the fundamental variables, qualified health
care provision (CHCI) was found to be strongly correlated to poverty and mother·s
autonomy as proxied by the M/F ratio (Table 5). It hence seems that autonomous mothers
are more capable of ensuring that their children are vaccinated and receive professional care
when sick. This can be expected to result in lower frequency of prolonged ill health and less
retarded skeletal growth (stunting). More autonomous mothers are also likely to be able to
feed their offspring more varied and micro-nutrient dense (but comparatively expensive)
food, which is a necessary (but not sufficient) precondition for normal (genetic potential)
growth in infants and young children.27

c
c
c c
The Ásian enigma hypothesesµ have only partially been supported by the findings in this
study. In the OLS and IV panel regressions the proxy for women·s autonomy, the M/F ratio,
came out significant for stunting, but not in the regression for underweight (Table 1). In the
first difference regressions, the autonomy variable turned out insignificant and/or not
robust in all regressions (Table 3). The latter result is not totally surprising since the changes
in child underweight have been very small over time.
In the OLS panel regressions for underweight on confounding variables, however,maternal
weight failure (BMI<18.5) was highly significant, corroborating another énigmaµ
hypothesis, i.e. that malnutrition tends to be transmitted over generatio ns (Table 4).

,)( &7()+
In the policy-focused literature on child malnutrition, a distinction is usually made between
long-term and short-term interventions. In this paper, we have focused mainly on factors
that are expected to improve child nutritional status in the long term: poverty reduction and
increases in female literacy and autonomy. It is now widely agreed, not only among
economists, but also nutritionists, as well as analysts from the international organisations,
that substantial poverty reduction is a necessary and important long-term prerequisite for
accomplishing more rapid alleviation of child malnutrition.28 It is notable that the first MDG
is to halve poverty ´hungerµ before the year 2015; the merging of these two objectives in
the same goal reflects a generally held perception that they are closely related. In the present
paper we have reported results that confirm this in the case of India.
If poverty is to be reduced more forcefully in India than in the recent past, future economic
growth has to encompass households in the lowest income quintiles to a larger extent. In the
macro-economic literature on India, there seems to be almost consensus on the required
strategy, at least in broad terms. The prime focus should be on increasing labour
productivity (and hence incomes) in agriculture, which still (2005) employs 54% of the
Indian labour force. This share has dropped by 6 p ercentage points only from 1993 and India
is probably one of very few countries in which employment in th e agricultural sector is still
growing in absolute numbers. The productivity and income gaps between the agricultural
and the industry and service sectors have grown rapidly since the early 1990s and are
estimated to be around five-fold in recent years (Bosworth et al 2008).
However, as labour productivity in agriculture increases, surplus labour has to be absorbed
in other rapidly growing labour-intensive sectors. The small-scale (rural) manufacturing

c
c
c c
sector should have good potential, given the abundance of low-skilled labour in India, but
has so far not been expanding very rapidly.29
As we have seen, female illiteracy is a major drag on the alleviation of child malnutrition
through a multitude of channels (Table 5). That female illiteracy is still close to 50% in rural
India suggests ample scope for improvement that our results suggest would have a
significant impact on child malnutrition. In this perspective, it is encouraging that female
secondary school enrolment in rural areas is on the increase in many states (but with some
notable exceptions). More education for males may also be helpful for eroding the
conservative values underlying the discrimination of women, as reflected in the high and
persistent M/F population ratios.
In most of the policy-oriented epidemiology-cum-nutrition literature, the focus is on

targeted interventions to alleviate child malnutrition in the short and medium terms.
There is a wealth of evaluations demonstrating high returns to such interventions in strictly
controlled experiments in small select communities (Allen and Gillespie 2001; Behrman et al
2004; Horton et al 2008; Lancet appendix 2008). There is a dearth of reliable evaluations of
large, scaled-up interventions in developing countries aimed at entire (sub)-populations.
The available studies find, however, that most such interventions have been poorly targeted
to the intended groups, e.g. malnourished children (Coady et al 2004; Lancet appendix
2008).116
Some of the results derived in the present paper suggest that targeted interventions to
children should have promising effects on the reduction of child malnutrition in India in the
not-so-long term. Increased maternal education on feeding practices and on the importance
of qualified health care provision when children are sick are two examples.
Much of this has been tried in India for many decades.
In India, the largest scaled-up program by far, is the Indian Child Distribution System
(ICDS), which on paper covers two-thirds of all villages in India. The program has recently
been evaluated by independent researchers, by at least three official Indian government
commissions, and by the World Bank.
116 In a recent study from Mexico, 5-10% of the children were found to be both stunted and
overweight, indicating micro-nutrient deficiency and overindulgence of calorie-rich staple
food (Fernald and Neufeld 2006).
c
c
c c
30 All found the ICDS in general to be underfinanced, ill targeted and inefficiently managed,
and hence to have little or no impact on children.
The federal Indian government has limited juridic al and financial power over the health-care
sector in the states, but provides the main funding for the ICDS and a number of other
programs aimed at improving child and maternal welfare.
More generous federal government financial support for this program is underway since
2004/05 (GOI 2006).117
This is a promising start, but as the evaluations sh ow, what is also required is an efficiency-
enhancing overhaul of the program, which has yet to materialise. As noted in a recent
official evaluation: Áfter 30 years of rich experience in the programmatic perspective, a
paradigm shift is required to reform the ICDS in respect of overall programme management
for a faster and sustained achievement of child and women nutritional goalsµ (GOI 2007c).
There is no lack of suggestions for how to enhance the efficiency of the ICDS (e.g. Levinsson
et al 2005) and there are local success stories (Tarozzi 2005) that may be possible to emulate
in other states.
Another option is to replace (or supplement) the ICDS and other defunct existing programs
with some form of conditional cash transfer scheme of the type that Mexico
(Progresa/opportunidad) and Brazil (Bolsa famiglia) have initiated.
These are among the few large-scale child programs in developing countries that have been
efficient and effective according to a number of evaluations. 118
It may be that conditional cash-transfer programs work well in Latin American countries
because child malnutrition is heavily concentrated geographically and to the poorest
households (Svedberg 2007). In India, where malnutrition is widespread (cf. figures 3 and 4),
117 See, for instance, the two series of articles on Ćhild development in developing
countriesµ and ´Maternal and child under nutritionµ, published in The Lancet in early 2008.
Also see World Bank 2004.
118 For more detailed quantitative analyses along these lines, see Datt and Ravallion 2002;
Foster and Rosenzweig 2004; Loayza and Raddatz 2006; Bosworth et al 2007; Kraay 2007;
Honorati and Mengistae 2007; Mitra and Ural 2007; World Bank 2008; Lal 2008; Panagarjya
2008; Subramanian 2008.

c
c
c c
targeting could be more problematic and costly (Adato and Hoddinott 2007). Nevertheless,
cash programs ought to be considered, but have so far not been widely discussed in India.
The recent focus seems to have been on expanding and rejuvenating the hitherto dismal
(Gahia 1996) National Rural Employment Guarantee Scheme.
In sum, the shining overall economic performance of the Indian economy since the early
1990s has undoubtedly left large population segments in the shade, as reflected in slow
declines of poverty and child malnutrition. The most promising route ahead is a
combination of an overall long-term economic growth strategy that is more inclusive of the
poor population groups paired with improved targeted interventions for the alleviation of
child malnutrition in the shorter term. Whatever strategies and programs that are opted for,
however, the prospects for success depend on financial funding, commitment and
operational efficiency >and hence on political priorities.119
119 Lonchin et al 2006; Kochar 2005; GOI 2005, 2006, 2007c; Granolati 2006; World Bank
2004, 2006.

c
c
c c

c
c
c c

c
c
c c

c
c
c c

c
c
c c
c
c

c
c
c c
$&&
1.c The facts: Child malnutrition in India
India is home to 40 percent of the world·s malnourished children and 35 percent of the
developing world· slow-birth-weight infants; every year 2.5 million children die in India,
accounting for one in five deaths in the world. More than half of these deaths could be
prevented if children were well nourished. India·s progress in reducing child malnutrition
has been slow. The prevalence of child malnutrition in India deviates further from the
expected level at the country·s per capita income than in any other large developing country.
2. The challenge: Accelerating progress in reducing child malnutrition in India

India has many nutrition and social safety net programs, some of which (such as Integrated
Child Development Services [ICDS] and the Public Distribution System [PDS]) have had
success in several states in addressing the needs of poor households. All of these programs
have potential, but they do not form a comprehensive nutrition strategy, and they have not
addressed the nutrition problem effectively so far.
3. Strategic choices for improved child nutrition India lacks a comprehensive nutrition
strategy. Various choices for nutrition strategies can be considered. A review of some of the
more successful country experiences suggests that all of them implemented complex,
multisectoral actions with more or less emphasis on service-oriented nutrition policies (as in
Indonesia),incentive-oriented nutrition policies linked to community or household
participation and performance (as in Mexico), or mobilization-oriented nutrition policies (as
in Thailand). These choices are not mutually exclusive. India now has the opportunity to
´leapfrogµ toward innovative nutritional improvement based on the experiences of other
countries and on experiences within India itself.
4. Cooperation for policy actions: To accelerate progress in reducing child malnutrition,

India should focus on the following four cross-cutting strategic approaches:
ensuring that economic growth and poverty reduction policies reach the poor;
' redesigning nutrition and health policies and programs by drawing on science and
technology for nutritional improvement, strengthening their implementation, and increasing
their coverage;

c
c
c c
increasing investments and actions in nutrition services for communities with th e highest
concentration of poor; and
3 focusing programs on girls· and women·s health and nutrition.

The main objective of this study has been to exp lain why the reduction in child m alnutrition
has been relatively small despite the impressive overall performance of the Indian economy
since the early 1990s. Although the results are only indicative, we have found that while
poverty reduction has a significant impact on the alleviation of child malnutrition in India
However, poverty decline has been modest despite high aggregate growth in the economy.
This, in turn, is at least partly a consequence of slow growth of household real consumption
expenditures among the poorest quintiles that are predominately employed in the
agricultural sector. In this sector, factor (labour), productivity growth has been much slower
than in the rest of the Indian economy and even declined since the late 1990s
Female illiteracy was found to be a strong determinant of child malnutrition, which is in line
with results in earlier related literature. In all-India, female illiteracy has declined notably
since the early 1990s, from 55% to 39% in 2005. Masked behind these averages, however, is
the fact that female illiteracy fell less in the rural areas of the most populous states, with the
initially highest prevalence of child malnutrition.
In these states, Bihar, Madhya Pradesh, Rajasthan and Uttar Pradesh, rural female illiteracy
was still well above 50% in 2005 and the rural population accounts for 75-87% of the total in
these states.26 Overall improvements in female literacy has helped bring down child
malnutrition according to the results reported here, but in rural India, female illiteracy is still
more than twice as high as in urban settings (46 vs. 20% in 2005).
The third fundamental explanatory variable for child malnutrition in this study, the M/F
population ratio, used as a proxy for women·s autonomy, was found to have a significant
impact on child stunting in the panel and first -difference regressions, but not on
underweight. In all-India, the M/F ratio has changed only marginally over time, from 26 In
the 2001 census, the three (pre-secession) states had a joint population of 374 million, or
36.4% of the total in India. The population in Uttar Pradesh, at 175 million, exceeded by far
the entire population in Pakistan. A recent nationally representative estimate of the M/F
ratio at birth found it to be 111.2, reflecting mainly gender selective abortions . This signals
that the gender bias in India is not about to erode in the near future. In addition to
estimating the relative strength of fundamental variables behind child malnutrition, we have
examined two pathways through which children·s nutritional status are assumed to be
c
c
c c
affected. The first is the link from mother to child nutriti onal status. Child underweight (but
not stunting) was found to be highly correlated to underweight among mothers. This is in
line with the world-wide observation that malnutrition in expecting mothers is a strong
predictor of LBW and subsequent underweight in infants and young children.
Unfortunately, we were not able to test the LBW link directly due to the unavailability of
data on birth weights. Underweight in mothers (BMI<18.5) themselves was found to be
significantly correlated to poverty and own illiteracy (but not the M/F ratio), the same
fundamental variables behind child underweight. The second pathway focused on the link
from female illiteracy, women·s turnout in state election and state government expenditures
on health care.
The variable women/men turnout ratio was found to be strongly associated with women
literacy, but also with poverty and total turnout (female and male). A high women/men
turnout ratio was identified as a highly significant determinant of state health expenditures.
In this regression, the total turnout was significant, but carried a negative sign. The other
control variable, NSDP/C, turned out insignificant. In the regression aimed at finding out
whether qualified child health care, as proxied by the CHCI, depends on SGHE/C, t his was
confirmed. Finally, it was found that child stunting (and less so underweight) is strongly
associated with the provision of health care as measured by the CHCI. In the regressions for
confounding variables on the fundamental variables, qualified he alth care provision (CHCI)
was found to be strongly correlated to poverty and mother·s autonomy as proxied by the
M/F ratio.
This can be expected to result in lower frequency of prolonged ill health and less retarded
skeletal growth (stunting). More autonomous mothers are also likely to be able to feed their
offspring more varied and micro-nutrient dense (but comparatively expensive) food, which
is a necessary (but not sufficient) precondition for normal (genetic potential) growth in
infants and young children.27 The Ásian enigma hypothesesµ have only partially been
supported by the findings in this study. In the OLS and IV panel regressions the proxy for
women·s autonomy, the M/F ratio, came out significant for stunting, but not in the
regression for underweight . In the first difference regressions, the autonomy variable
turned out insignificant and/or not robust in all regressions. The latter result is not totally
surprising since the changes in child underweight have been very small over time. In the
OLS panel regressions for underweight on confounding variables, however, maternal
weight failure (BMI<18.5) was highly significant, corroborating another énigmaµ
c
c
c c
hypothesis, i.e. that malnutrition tends to be transmitted over generations In the policy-
focused literature on child malnutrition, a distinction is usually made between long-term
and short-term interventions.
In this paper, we have focused mainly on factors that are expected to improve child
nutritional status in the long term: poverty reduction and increases in female literacy and
autonomy. It is now widely agreed, not only among economists, but also nutritionists, as
well as analysts from the international organisations, that substantial poverty reduction is a
necessary and important long-term prerequisite for accomplishing more rapid alleviation of
child malnutrition.28 It is notable that the first MDG is to halve poverty ´hungerµ before
the year 2015; the merging of these two objectives in the same goal reflects a generally held
perception that they are closely related. In the present paper we have reported results that
confirm this in the case of India.
If poverty is to be reduced more forcefully in India than in the recent past, future economic
growth has to encompass households in the lowest income quintiles to a larger extent. In the
macro-economic literature on India, there seems to be almost consensus on the required
strategy, at least in broad terms. The prime focus should be on increasing labour
productivity (and hence incomes) in agriculture, which still (2005) employs 54% of the
Indian labour force. This share has dropped by 6 percentage points only from 1993 and India
is probably one of very few countries in which employment in the agricultural sector is still
growing in absolute numbers. The productivity and income gaps between the agricultural
and the industry and service sectors have grown rapidly since the early 1990s and are
estimated to be around five-fold in recent years (Bosworth et al 2008). However, as labour
productivity in agriculture increases, surplus labour has to be absorbed in other rapidly
growing labour-intensive sectors. The small-scale (rural) manufacturing sector should have
good potential, given the abundance of low-skilled labour in India, but has so far not been
expanding very rapidly.29
As we have seen, female illiteracy is a major drag on the alleviation of child malnutrition
through a multitude of channels. That female illiteracy is still close to 50% in rural India
suggests ample scope for improvement that our results suggest would have a significant
impact on child malnutrition.
In this perspective, it is encouraging that female secondary school enrolment in rural areas is
on the increase in many states (but with some notable exceptions). More education for males

c
c
c c
may also be helpful for eroding the conservative values underlying the discrimination of
women, as reflected in the high and persistent M/F population ratios.
In most of the policy-oriented epidemiology-cum-nutrition literature, the focus is on
targeted interventions to alleviate child malnutrition in the short and medium terms.
There is a wealth of evaluations demonstrating high returns to such interventions in strictly
controlled experiments in small select communities. There is a dearth of reliable evaluations
of large, scaled-up interventions in developing countries aimed at 28 See, for instance, the
two series of articles on Ćhild development in developing countriesµ and ´Maternal and
child under nutritionµ, published in The Lancet in early 2008.
The available studies find, however, that most such interventions have been poorly targeted
to the intended groups, e.g. malnourished children .
Some of the results derived in the present paper suggest that targeted interventions
to children should have promising effects on the reduction of child malnutrition in India in
the not-so-long term. Increased maternal education on feeding practices and on the
importance of qualified health care provision when children are sick are two examples.
Much of this has been tried in India for many decades.
In India, the largest scaled-up program by far, is the Indian Child Distribution System
(ICDS), which on paper covers two-thirds of all villages in India. The program has recently
been evaluated by independent researchers, by at least three official Indian government
commissions, and by the World Bank.30 All found the ICDS in general to be underfinanced,
ill targeted and inefficiently managed, and hence to have little or no impact on children.
The federal Indian government has limited juridical and financial power over the health-care
sector in the states, but provides the main funding for the ICDS and a number of other
programs aimed at improving child and maternal welfare. More generous federal
government financial support for this program is underway since 2004/05 (GOI 2006).
This is a promising start, but as the evaluations sh ow, what is also required is an efficiency-
enhancing overhaul of the program, which has yet to materialise.
As noted in a recent official evaluation: Áfter 30 years of rich experience in the
programmatic perspective, a paradigm shift is required to reform the ICDS in respect of
overall programmed management for a faster and sustained achievement of child and
women nutritional goalsµ. There is no lack of suggestions for how to enhance the efficiency

c
c
c c
of the ICDS and there are local success stories that may be possible to emulate in other
states.
Another option is to replace (or supplement) the ICDS and other defunct existing Programs
with some form of conditional cash transfer scheme of the type that Mexico
These are among the few large-scale child programs in developing countries that have been
efficient and effective according to a number of evaluations.31 It may be that conditional
cash-transfer programs work well in Latin American countries because child malnutrition is
heavily concentrated geographically and to the poorest households.
In India, where malnutrition is widespread , targeting could be more problematic and costly.
Nevertheless, cash programs ought to be considered, but have so far not been widely
discussed in India. The recent focus seems to have been on expanding and rejuvenating the
hitherto dismal
)+($(&7() &+$+&
In sum, the shining overall economic performance of the Indian economy since the early
1990s has undoubtedly left large population segments in the shade, as reflected in slow
declines of poverty and child malnutrition. The most promising route ahead is a
combination of an overall long-term economic growth strategy that is more inclusive of the
poor population groups paired with improved targeted interventions for the alleviation of
child malnutrition in the shorter term. Whatever strategies and programs that are opted for,
however, the prospects for success depend on financial funding, commitment and
operational efficiency >and hence on political priorities.
Malnutrition, defined as underweight, is a serious public -health problem that has been l
inked to a substantial increase in the risk of mortality and morbidity. Women and young
children bear the brunt of the disease burden associated with malnutrition. In Africa and
south Asia, 2751% of women of reproductive age are underweight (ACC/SCN, 2000), and
it is predicted that about 130 million children will be underweight in 2005 (21% of all
children) (de Onis et al., 2004a). Many of the 30 million low-birth-weight babies born
annually (23.8% of all births) face severe short-term and long-term health consequences.
In this guide we outline a method for estimating the disease burden at national or local level
that is associated with maternal and child malnutrition. The goal is to help policymakers and

c
c
c c
others quantify the increased risk associated with malnutrition, in terms of attributable
mortality and morbidity, at country or local levels. The estimates will allow policy -makers to
compare the disease burden of malnutrition for different countries, or regions within
countries, and enable resources to be deployed more effectively.
Repeated assessments will also allow trends to be monitored and the impact ofinterventions
to be evaluated. To quantify the disease burden, population attributable fractions are
derived from the assessed exposure (malnutrition) and from the relative risk estimates of
disease and death associated with malnutrition. The level of malnutrition in the population
groups is assessed by anthropometry (i.e. measurements of body size and composition),
using as indicators low birth weight in newborns, low weight -for-age in preschool children,
and low body mass index in women. Relative risk estimates for diarrhea, malaria, measles,
acute respiratory infections and other infectious diseases are based on a meta-analysis that
was part of a global comparative risk assessment project conducted by the World Health
Organization (WHO) and its partners. Checklists for collecting and analyzing data are also
suggested, and a step-by-step example of how to quantify the health impact associated with
malnutrition is given sub region. Estimates of the disease burden of malnutrition give
policy-makers an indication of the burden that could be avoided if malnutrition were to be
eliminated. Disaggregated estimates (e.g. by age, sex, degree of malnutrition) can also help
policy-makers identify the segments of a population most at risk, such as women and
children, and direct resources where they will have the greatest effect. Although it is difficult
to assess the avoidable burden because of the uncertainties around estimates of risk factors
and disease burdens, the importance of the avoidable burden for policy -making justifies the
effort (WHO, 2002). Reducing hunger and malnutrition, endemic in many parts of
developing world, is intrinsically important. But so too are many of the other challenges
considered at this conference.
Our focus here has been on the economic case for investing in activities that reduce hunger
and malnutrition. We have focused on four opportunities
Reducing the Prevalence of Low Birth weight
Infant and Child Nutrition and Exclusive Breastfeeding Promotion
Reducing the Prevalence of Iron Deficiency Anemia and Vitamin A, Iodine and Zinc
Deficiencies
Investment in Technology in Developing Country Agriculture

c
c
c c
While the opportunities described here represent our current view of the most promising
approaches several other opportunities, in part covered by other Challenge Papers, have
important implications for reducing hunger and malnutrition in that successes in these other
challenges will affect nutrition and, conversely improvements in nutrition will influence the
other challenges
For example: - Women·s education and status. Numerous studies show strong correlations
between maternal education and reductions in under nutrition amongst pre -school children.
While some of this work is subject to caveats (for example, maternal education may be
correlated with unobserved family background characteristics), progress on the challenge
relating to lack of education is likely to produce benefits in terms of reduced malnutrition. -
Addressing infectious diseases such as malaria and the HIV/AIDS pandemic. For example,
HIV/AIDs increases hunger and malnutrition directly by reducing the income and food
security of affected households and by interfering with the intergenerational transmittal of
agricultural and other productive skills. In addition, young orphans and children with
chronically ill caregivers risk higher rates of malnutrition. HIV also imposes a dilemma in
assessing the increased risks of breastfeeding against the risks of not breastfeeding.
Conversely, nutrition affects HIV/AIDS since nutritional status is a major factor influencing
a person·s risk of infection.
Well-nourished HIV positive individuals live longer; and respond more positively to
treatment. Moreover, the efficacy of certain anti-retroviral drugs is diminished when they
are not taken on a full stomach.
Improving infrastructure to reduce possibilities of famine or chronic hunger. Famines and

chronic under nutrition typically currently do not reflect food shortages in the aggregate so
much as inadequate access to food for poorer segments of the population ² either due to
short-run shocks or chronic conditions. Inadequate food access, in turn, reflects limited
purchasing power in the short run or longer-run, often exacerbated by food price shocks in
partially segmented markets.
Communication and transportation infrastructure investments can serves to lessen localized
prices shocks that may be an important factor in famines, and may lead to increased growth
and therefore lessened chronic under nutrition. These estimates only use a 3% discount rate.
Again assuming that the benefits are distributed evenly over the 8-25 year period, a 5%
discount rate produces a benefit-cost ratio of 8.5
c
c
c c
- Effective improvements in water and sanitation lead to reductions in diarrhea. As noted in

Opportunity 2, repeated diarrheal infections are correlated with growth faltering,
particularly for children less than three years of age.
- Trade barriers impose considerable costs on developing countries. The majority of hungry
and malnourished people in developing countries are poor. The majorities of poor people in
developing countries live in rural areas and depend directly or indirectly on agriculture for
their livelihoods. Changes in the returns to agriculture in developing countries, thus, may
have a major impact on hunger and malnutrition in developing countries through affecting
the income of the poor and through affecting the prices that the poor pay for basic staples
and other foods.
These changes, in turn, may have impacts throughout the lifecycle, from conception
onwards, since changes in the resources that households have and the prices that they face
may trigger nutritional changes over the lifecycle with effects such as are reviewed in the
last part of Section 2.
In the four opportunities that we have enumerated to directly address hunger and
malnutrition -subject to the caveats we enumerate - there are a range of considerable benefits
from: resource saving arising from reduced mortality; resource saving arising from reduced
morbidity; direct links between nutrition and physical productivity; and indirect links
between nutrition, cognitive development, schooling and productivity. While we have
drawn upon recent evidence of project impacts and costs we note that our main conclusions
about the high returns to investing in nutrition and in agricultural technology repeat general
observations that have been made earlier in the literature and yet, to a fair degree, the
potential investments remain to remain under-resourced. It suggest considerable possible
gains to be had in investing in these opportunities, in the sense of benefit-to-cost ratios
exceeding one or relatively high internal rates of return to investing in programs or policies
to reduce hunger and malnutrition ² in addition to the intrinsic welfare gains to the
individuals who would be effected directly by reduced hunger and malnutrition.
The gains appear to be particularly large for reducing micronutrient deficiencies in
populations in which prevalence·s are high. Moreover, the people who are likely to benefit
from these interventions tend to be relatively poor, so such investments are likely to have
important gains in terms of the objective of reducing poverty as well as in terms of
c
c
c c
increasing productivity. Finally, while the available studies generally do not distinguish well
between private and social rates of returns to these interventions, on the basis of limited
studies and casual observations it would appear that there are important aspects of the
potential gains that are social beyond the private gains due to externalities related to
contagious diseases and to education ² so there is likely to be a case for the use of some
public resources for such interventions on efficiency grounds in addition to the case on
poverty alleviation grounds.
It is of little value to assess child growth if the assessment is not followed-up by policy
action to improve the health and nutritional status of children. Policy actions are likely to
have the greatest impact on child malnutrition if they are directed at the early stages of child
development (i.e. pre-natal, infancy and early childhood) (Shrimpton, 2001). A good start in
life will pay off, both in terms of human capital and economic development. Interventions
that improve the physical growth and mental development of children will not only
decrease the prevalence of underweight, but also prevent its negative functional
consequences throughout life. Reducing malnutrition thus not only benefi ts child health and
development in the short term, it also promotes the future, long-term growth and economic
progress of the nation.
For these reasons, the nutritional status of children should be a primary indicator of
socioeconomic development, and good child nutritional status should be considered a
precondition for the long-term socioeconomic progress of a society. Children are the most
vulnerable members of society and to allow their development to be affected by poor
nutrition is to perpetuate the vicious cycle of poverty and malnutrition, and to waste human
potential. Governments will be unable to accelerate economic development over the long
term until their children are assured of optimal growth and development.
Malnutrition is an impediment to development, and its presence indicates that basic

physiological needs have not been met. What is observed as malnutrition is not only the
result of insufficient or inappropriate food, but also a consequence of other condit ions, such
as poor water supply and sanitation and a high prevalence of disease. Thus reversing the
procedure is complex, because many issues need to be addressed more or less
simultaneously. And every situation is different, so that there is no single solution for all.
There can only be general guidance on directions to pursue. Experience from lessons learnt
c
c
c c
shows that considerable time is needed to redress a situation (ten years and more), and that
a strong supportive political and policy environment remains crucial throughout the period.
There is no ´quick-fixµ to this problem. Once achieved, however, the effect is likely to
become permanent, offering a substantial return on investment.µ
Malnutrition occurs when the body does not get enough vitamins, minerals, and other
nutrients it needs to maintain healthy tissues and organ function. Both undernourished or
over-nourished people can suffer from malnutrition
Under nutrition is a consequence of consuming too few essential nutrients or using or

excreting them more rapidly than they can be replaced.
Infants, young children, and teenagers need additional nutrients. So do women who are
pregnant or breastfeeding. Nutrient loss can be accelerated by diarrhea, excessive sweating,
heavy bleeding (hemorrhage), or kidney failure. Nutrient intake can be restricted by age-
related illnesses and conditions, excessive dieting, severe injury, serious illness, a lengthy
hospitalization, or substance abuse.
The leading cause of death in children in developing countries is protein-energy

malnutrition. This type of malnutrition results from inadequate intake of calories from
proteins, vitamins, and minerals. Children who are already undernourished can suffer from
protein-energy malnutrition when rapid growth, infection, or disease increases the need for
protein and essential minerals.
In the United States, nutritional deficiencies generally have been replaced by dietary
imbalances or excesses associated with many of the leading causes of death and disability.
Over nutrition results from eating too much, eating too many of the wrong things, not
exercising enough, or taking too many vitamins or other dietary replacements.
Risk of over nutrition is also increased by being more than 20% overweight, consuming a
diet high in fat and salt, and taking high doses of:
Ôc Nicotinic acid (niacin) to lower elevated cholesterol levels

Ôc Vitamin B 6 to relieve premenstrual syndrome
Ôc Vitamin A to clear up skin problems
Ôc Iron or other trace minerals not prescribed by a doctor.

c
c
c c
Nutritional disorders can affect any system in the body and the senses of sight, taste, and
smell. Malnutrition begins with changes in nutrient levels in blood and tissues. Alterations
in enzyme levels, tissue abnormalities, and organ malfunction may be followed by illness
and death.
Poverty and lack of food are the primary reasons why malnutrition occurs in the United
States. Ten percent of all low income households members do not always have enough
healthful food to eat, and malnutrition affects one in four elderly Americans. Protein-energy
malnutrition occurs in 50% of surgical patients and in 48% of all other hospital patients.
There is an increased risk of malnutrition associated with chronic diseases, especially disease
of the intestinal tract, kidneys, and liver. Patients with chronic diseases like cancer, AIDS,
and intestinal disorders may lose weight rapidly and become susceptible to
undernourishment because they cannot absorb valuable vitamins, calories, and iron.
People with drug or alcohol dependencies are also at increased risk of malnutrition. These
people tend to maintain inadequate diets for long periods of time, and their ability to absorb
nutrients is impaired by the alcohol or drug's effect on body tissues, particularly the liver,
pancreas, and brain.
Unintentionally losing 10 pounds or more may be a sign of malnutrition. People who are
malnourished may be skinny or bloated. Their skin is pale, thick, dry, and bruises easily.
Rashes and changes in pigmentation are common.
Hair is thin, tightly curled, and pulls out ea sily. Joints ache and bones are soft and tender.
The gums bleed. The tongue may be swollen or shriveled and cracked. Visual disturbances
include night blindness and increased sensitivity to light and glare.
Other symptoms of malnutrition include:
Ôc Anemia
Ôc Diarrhea
Ôc Disorientation
Ôc Goiter (enlarged thyroid gland)
Ôc Loss of reflexes and lack of coordination
Ôc Muscle twitches
Ôc Scaling and cracking of the lips and mouth.

c
c
c c
Malnourished children may be short for their age, thin, listless, and have weakened immune
systems.
Overall appearance, behavior, body-fat distribution, and organ function can alert a family
physician, internist, or nutrition specialist to the presence of malnutrition. Patients may be
asked to record what they eat during a specific period. X rays can determine bone density
and reveal gastrointestinal disturbances, and heart and lung damage.
Blood and urine tests are used to measure levels of vitamins, minerals, and waste
products. Nutritional status can also be determined by:
Ôc Comparing a patient's weight to standardized charts

Ôc Calculating body mass index (BMI) according to a formula that divides height into
weight
Ôc Measuring skin-fold thickness or the circumference of the upper arm.
Normalizing nutritional status starts with a nutritional assessment. This process enables a
clinical nutritionist or registered dietician to confirm the presence of malnutrition, assess the
effects of the disorder, and formulate diets that will restore adequate nutrition.
Patients who cannot or will not eat, or who are unable to absorb nutrients taken by mouth,
may be fed intravenously (parenteral nutrition) or through a tube inserted into the
gastrointestinal (GI) tract (enteral nutrition).
Tube feeding is often used to provide nutrients to patients who have suffered burns or who
have inflammatory bowel disease. In this procedure, a thin tube is inserted through the nose
and carefully guiding along the throat until it reaches the stomach or small intestine. If long -
term tube feeding is necessary, the tube may be placed directly into the stomach or small
intestine through an incision in the abdomen.
Tube feeding cannot always deliver adequate nutrients to patients who:
Ôc Are severely malnourished

Ôc Require surgery
Ôc Are undergoing chemotherapy or radiation treatments
Ôc Have been seriously burned
Ôc Have persistent diarrhea or vomiting

c
c
c c
Ôc Whose gastrointestinal tract is paralyzed.
Intravenous feeding can supply some or all of the nutrients these patients need.
Up to 10% of a person's body weight can be lost without side effects, but if more than 40 % is
lost, the situation is almost always fatal. Death usually results from heart failure, electrolyte
imbalance, or low body temperature. Patients with semi consciousness, persistent diarrhea,
jaundice, or low blood sodium levels have a poorer prognosis.
Some children with protein-energy malnutrition recover completely. Others have many
health problems throughout life, including mental retardation and the inability to absorb
nutrients through the intestinal tract. Prognosis for patients with malnutrition seems to be
dependent on the patient's age and the length and severity of the malnutrition, with young
children and the elderly having the highest rate of long -term complications and death.
Breastfeeding a baby for at least six months is considered the best way to prevent early-
childhood malnutrition. The United States Department of Agriculture and Health and
Human Service recommend that all Americans over the age of two:
Ôc Consume plenty of fruits, grains, and vegetables

Ôc Eat a variety of foods that are low in fats and cholesterols and contain only moderate
amounts of salt, sugars, and sodium
Ôc Engage in moderate physical activity for at least 30 minutes, at least several times a
week
Ôc Achieve or maintain their ideal weight
Ôc Use alcohol sparingly or avoid it altogether.
Every patient admitted to a hospital should be screened for the presence of illnesses and
conditions that could lead to protein-energy malnutrition. Patients with higher -than-average
risk for malnutrition should be more closely assessed and reevaluated often during long-
term hospitalization or nursing-home care.
We believe that minimizing malnutrition in developing countries such as India consists of

undertaking short term procedures such as donating money, medical supplies and sending
doctors to help the victims. Though this may sound like a good idea, we learnt from a guest
speaker that you shouldn·t give someone a fish, but rather teach them how to catch it, so our
long term proposal is to try to reduce malnutrition by first diminishing the problem of

c
c
c c
overpopulation. We plan to do this by using Thomas Malthus·s preventative checks theory

of lowering the birth rate.
Currently, India is not a sustainable country, as it is overpopulated and food distribution is

very unequal between the upper class and l ower class. To create a sustainable future for our
blue planet, India should take into account the theories of Thomas Malthus, and control their
birth rate in order to equalize food distribution and therefore minimize malnutrition. If
India·s population is balanced, everyone will get enough food and no person will be left
malnourished, which results in a sustainable future for our blue planet.

c
c
c c
PART C
METHODOLOGY

c
c
c c
"+

We have used two research designs in our project.
2. c Exploratory Research
3. c Descriptive Research
#7()) : Exploratory research is a type of research conducted because

a problem has not been clearly defined. Exploratory research helps determine the
best research design, data collection method and selection of subjects. Given its
fundamental nature, exploratory research often concludes that a perceived problem
does not actually exist. In our case we first wanted to make certain aspects of the
study clear before actually describing the problem. The primary research too has
helped us achieve some clarity.
7%: This research is the most commonly used and the basic reason
for carrying out descriptive research is to identify the cause of something that is
happening. Descriptive research, also known as statistical research, describes data
and characteristics about the population or phenomenon being studied. Descriptive
research answers the questions who, what, where, when and how. The description is
used for frequencies, averages and other statistical calculations. Often the best
approach, prior to writing descriptive research, is to conduct a survey investigation.
Qualitative research often has the aim of description and researchers may follow-up
with examinations of why the observations exist and what the implications of the
findings are.
We have used this research along with exploratory research. For carrying out
descriptive research we have used a questionnaire, this questionnaire would be
administered to 90 respondents and this would help us to identify the various factor
which will help us to study the malnutrition.
,m ,
Distribution :
Strata Possible Respondents
Children 35
Students 50
Working Class 25
House-Wives 25
Conduct Field-work :
The data will be collected via Questionnaires
Determine Sample Size:
The Sampling unit of our student will include students, working class, housewives
(age group of 18-40) from lower middle and upper middle class. The sample size will
consist of approximately 100 sampling units.
+?7*=7@*;A@

c
c
c c
Where,
,denotes probability of success = 0.2
*=7@denotes probability of failure = 0.8
;denotes standard normal variate = 1.96 (As Confidence Interval is 95%)
denotes standard error = 8%
+denotes sample size
Therefore
n=(0.2)*(0.8)*(1.96/0.08)2 = 0.16 *631.42
n=100
Selecting an actual sample unit :

The actual sample units would consist of students present in colleges between Mira
Road and Andheri, the working class and the housing²wives living in the area.
Target Population :
The Target population would consist of children , students belonging to college
going onwards category. Other than that Working class and house-wives in the age
group of around 35 would also be considered.
Sampling Frame:
The Sampling Frame would consist of children, college going students, Working
class, and house-wives residing/working/studying in colleges between Mira Road
and Andheri.
Determination of Sampling Method :

The sampling method used for this project is 7)''( &7(+" &)3 .The
main reason for using this method is that it is possible to pre -specify every potential
sample of a given size that could be drawn from the population. Here the sampling
units are selected by chance.
Selecting a Sample Unit:
The procedure for selecting a sample units is *3 sampling. The population
would be divided into mutually exclusive and collectively exhaustive strata·s
depending on age and working status (working/non-working). Then the elements
from each strata would be selected by a random procedure, mainly simple random
sampling.

c
c
c c
m
)(()+)3
Method Selected- Survey method

Reasons:
- Our topic is about general public and we have to find out the response of thiers
towards various questions based on their eating habits. Hence we use survey
method which can capture a wide variety of information.
- Also it is relatively easy to administer.
c
')&&$+)+77)27)+3+
a. c Personal Interview
b. c E-mail
c. c Telephonic Interview
)(()+m+$&+
Questionnaire

c
c
c c
.m
The main data analysis tools used will be
-
The process includes the review of the data to ensure maximum accuracy and
unambiguity. Careful editing in the early process of data collection will be subject to
misunderstanding of instructions, errors in recording and other problems at a stage
when it is still possible to eliminate them from the later stages of the study.
! -
The process includes careful interpretation and good judgment of the data to ensure
that the meaning of the responses and the meaning of the category are consistently
and uniformly matched.
The data will be analyzed using the following statistical tools
i.c Correlation and regression.
ii. c Time-series
iii. c Hypothesis testing will also be undertaken.
To Aid in the process of Data Analysis we propose to use the following statistical
soft-wares:
i.c SPSS
ii. c Microsoft Excel
$( will be depicted through:

i.c Graphs: Bar diagrams
ii. c Pie Charts
iii. c Tables: Presentation of both qualitative and quantitative data

c
c
c c
+(

c What is your Occupation?
Chart 6.1
c
c
c
G c c c
c
c
c
c
c
c
O
c
c c c
c
Interpretations: The sample group of our survey consisted maximum number of

Working Class followed by the students.
c At what age did your child start weaning?
Chart 6.2
c
c
c
G c c c
c
c
c c
c
c
c
c cc cc ccc
c c
Interpretations: Maximum number of respondents children started weaning in the

Age Group of 1- 1.5 yrs

c
c
c c
c What is your educational level?

Chart 6.3
c
c
c
c
c
G c c c
c
c
c
c
c c
c G c c Gc c
c c
c c c
c
Interpretations: The respondents mainly belong to the post graduate group
0c What is the level of your family income?
Chart 6.4
c
c
c
c
c
G c c
c
c c
c c
c c
c
c
c c c c
c c c c c
c
c
Interpretations : Majority of respondents family income lies between Rs 50000-

100000

c
c
c c
4c What is your height?

Chart 6.5
c
c
c
c
G c c c
c
c c
c
c c
c
cc cc c cc
Interpretations: Among all our responders most of the people height is between 4-
5ft.

Îc What is your weight?
Chart 6.6
c
c
c
c
c
G c c
c c
c
c
c
c c
c
cc cc c c
Interpretations: Maximum numbers of our respondents are in the category of 40 -50

kgs

c
c
c c
åc How many meals do you have in a day?

Chart 6.7
c
c

c
c
c
G c c
c
c c
c c
c
c
c
c c c c c
c c c
c c
c
Interpretation: Most of the respondents have 3 meals a day
c How often do you go for health checkup?

Chart 6.8
c
c
c
c
c
G c c c
c
c c
c c
c c
c
c
,
c c c Occc c
c cc c
c
Interpretation: Majority of the people go for monthly checkup

c
c
c c
c When members of your family get sick where do u generally go for treatment?
Chart 6.9
c
c
c
c
c
G c c
c
c
c
c
c c
c
c
c
Gc G c c c c
Interpretation: Most people go to private dispensary

c
c
c c
mmm
1. c &)++: We had time constraints as being students; we were

involved in other assignments as well. Due to which we could not select a
larger sample size.This was the major constraint. Time for study on the project
is only 3 months
2. c 5$3")++: We also faced a budget constraint as this research is not

sponsored by any organization & is meant for academic purposes.
3. c )"7()++: We conducted the research only in Mumbai. This

also limited to administer the research to a larger survey.The results therefore
are confined to this area only and need not necessarily be applicable to other
areas.
4.c +)()" )++ The non-availability of the software for data analysis also
proved to be a limiting factor.
5. c 53 2 Respondents might have some biased views and this could
have affected the findings to a certain extent.
6. c )$7/The group members was only 4.
c
c
c
c
c
c
c
c
c
c
c
c

c
c
c c
PART D
FINDINGS AND
CONCLUSIONS

c
c
c c
c
c
The conclusion that could be drawn from the intensive research is that the
malnutrition rate in the population of Mumbai is not increasing major ly. The reasons
for the same are presented below:
Ôc A major portion of the urban population of Mumbai is health conscious. Most
of them carry home-made food at work and eat 3 meals per day
Ôc Even if they realize they are underweight, they follow a diet regime to
become healthy.
Ôc Irrespective of the nature of the job, the busy lifestyle in Mumbai helps to
keep most of them physically fit.
Ôc Also, Maximum number of respondents children started weaning in the Age Group
of 1- 1.5 Years.
Ôc Also, people are educated and health conscious and go for monthly check ups
regularly.
Ôc If they fall sick, they prefer to go to well established hygienic private

dispensaries.
Interpretations: The sample group of our survey consisted maximum number of

Working Class followed by the students.

c
c
c c
6, , c
,
m m
mm m
m m

$'&3)

m8.!6
)$7 )
%-5

)$7&'

%
%"2(00
2"2( 4
(31++44

#$%$&&

c
c
c c
The main purpose of conducting this research is to study and analyze the causes for
the increasing rate of malnutrition in India. This study will help us understand the
lifestyle of the people in rural & urban areas & also help to understand the degree of
awareness among people about malnutrition. In order to achieve our desired
objectives, we have collected data from the government records and social human
rights association, etc.
This study gives a bird·s eye view of the malnutrition statistics & trends prevalent
among the Indian population (both rural & urban). To accomplish the task, a survey
will be carried in and around Mumbai city with a sample of 100 respondents that
will be on a random. A structured questionnaire is prepared and will be
administered and the data so collected will be analyzed both by percentages and
statistical methods. Extensive use of Microsoft Excel software will be incorporated
for getting the analytical report.
The interviews will be conducted and data will be collected accordingly. Standard
editing and coding procedures would be utilized to ensure maximum accuracy and
unambiguity. This includes careful interpretation and good judgment of the data.
Hypothesis testing will also be undertaken. Simple tabulation will be utilized to
analyse the data.
A written report will be prepared and the presentation of the findings will be made.
A schedule of the research program as prepared will be followed so as to complete
the research on time and in budgeted limits.

c
c
c c
, 5
To study malnutrition

6 58m
c
c
c c

1.c To study the effect of lower income levels on malnutrition rate in India.
2.c To study the effect of inflation on malnutrition rate in India
3. c To study the effect of poor eating habits on malnutrition rate in India.

4. c To study the effect of inefficient Public Distribution System on malnutrition
rate in India
5. c To study the effect of inadequate storage facilities on malnutrition rate in
India

m,
.
c
c
c c

Ôc Malnutrition has become one of the most serious public health problems of
the 21st century with rates of adult and childhood malnutrition increasing. It
is one of the leading preventable causes of death worldwide.
Ôc This research will help us study the various aspects of malnutrition i.e. the
current malnutrition rate in the city, the various factors that contribute to
malnutrition and the impact of these factors on the p opulation.
Ôc This study will help us understand the lifestyle of the people in India and also
help to understand the degree of awareness among population towards
malnutrition.


c
c
c c
` `c c
c
We have used two research designs in our project.
1.c xploratory Research
2.c Descriptive Research
c`
: xploratory research is a type of research conducted because
a problem has not been clearly defined. xploratory research helps determine the
best research designÿ data collection method and selection of subjects. Given its
fundamental natureÿ exploratory research often concludes that a perceived problem
does not actually exist. In our case we first wanted to make certain aspects of the
study clear before actually describing the problem. The primary research too has
helped us achieve some clarity.

c`
: This research is the most commonly used and the basic reason
for carrying out descriptive research is to identify the cause of something that is
happening. Descriptive researchÿ also known as statistical researchÿ describes data
and characteristics about the population or phenomenon being studied. Descriptive
research answers the questions whoÿ whatÿ whereÿ when and how. The description is
used for frequenciesÿ averages and other statistical calculations. Often the best
approachÿ prior to writing descriptive researchÿ is to conduct a survey investigation.
Qualitative research often has the aim of description and researchers may follow-up
with examinations of why the observations exist and what the implications of the
findings are.
We have used this research along with exploratory research. For carrying out
7descriptive research we have used a questionnaireÿ this questionnaire would be
administered to 90 respondents and this would help us to identify the various factor
which will help us to study the malnutrition

c
c
c c
.m
The main data analysis tools used will be
-
The process includes the review of the data to ensure maximum accuracy and
unambiguity. Careful editing in the early process of data collection will be subject to
misunderstanding of instructions, errors in recording and other problems at a stage
when it is still possible to eliminate them from the later stages of the study.
! -
The process includes careful interpretation and good judgment of the data to ensure
that the meaning of the responses and the meaning of the category are consistently
and uniformly matched.
The data will be analyzed using the following statistical tools
iv. c Correlation and regression.
v. c Time-series
vi. c Hypothesis testing will also be undertaken.
To Aid in the process of Data Analysis we propose to use the following statistical
soft-wares:
iii. c SPSS
iv. c Microsoft Excel
$( will be depicted through:

iv. c Graphs: Bar diagrams
v. c Pie Charts
vi. c Tables: Presentation of both qualitative and quantitative data
c
c
c
c
c
c
c
c
c
c
c


c
c
c c
6., 6
6. c H0 There is no significant effect of lower income levels on

malnutrition rate
H1 There is significant effect of lower income levels on
malnutrition rate
7. c H0 There is no significant effect of inflation on malnutrition rate.
H1 There is significant effect of inflation on malnutrition rate.
8. c H0 There is no significant effect of poor eating habits on

malnutrition rate.
H1 There is significant effect of poor eating habits on malnutrition
rate
9. c H0 There is no significant effect of inefficient Public Distribution

H1 There is significant effect of inefficient Public Distribution
10. cH0 There is no significant effect of inadequate storage facilities on

malnutrition rate
H1 There is no significant effect of inadequate storage facilities on
malnutrition rate

c
c
c c
,m ,
Distribution :
Strata Possible Respondents

Children 35
Students 50
Working Class 25
House-Wives 25
Conduct Field-work :
The data will be collected via Questionnaires
Determine Sample Size:
The Sampling unit of our student will include students, working class, housewives
(age group of 18-40) from lower middle and upper middle class. The sample size will
consist of approximately 100 sampling units.
+?7*=7@*;A@
Where,
,denotes probability of success = 0.2

*=7@denotes probability of failure = 0.8
;denotes standard normal variate = 1.96 (As Confidence Interval is 95%)
denotes standard error = 8%
+denotes sample size
Therefore
n=(0.2)*(0.8)*(1.96/0.08)2 = 0.16 *631.42
n=100
Selecting an actual sample unit :

c
c
c c
The actual sample units would consist of students present in colleges between Mira
Road and Andheri, the working class and the housing²wives living in the area.
Target Population :
The Target population would consist of children ,students belonging to college going
onwards category. Other than that Working class and house-wives in the age group
of around 35 would also be considered.
Sampling Frame:
The Sampling Frame would consist of children, college going students, Working
class, and house-wives residing/working/studying in colleges between Mira Road
and Andheri.
Determination of Sampling Method :
The sampling method used for this project is 7)''( &7(+" &)3 .The
main reason for using this method is that it is possible to pre -specify every potential
sample of a given size that could be drawn from the population. Here the sampling
units are selected by chance.
Selecting a Sample Unit:
The procedure for selecting a sample units is *3 sampling. The population
would be divided into mutually exclusive and collectively exhaustive strata·s
depending on age and working status (working/non-working). Then the elements
from each strata would be selected by a random procedure, mainly simple random
sampling.

c
c
c c
> m
mm
6

Name Qualification
Nishi Avasthi B.Com
Ravi Agarawal B.Com
Swati Agrawal B.E
VaishaliRadhakrishanan B.Com
c
c
c
c c
m .5

% 3$( +3 ) ,$( &3
(
+) 3 3 )* )+
3
1 Problem 1st January 1 1 1 Printing 20 Computer
discovery printer
&definition
2 Problem 1st January 1 1 1 Printing 40 Computer
definition printer
3 Research 2nd 2 3 2 Internet 10 Computer
objective January- accessing printer
3rd January
4 Research 4th 4 12 9 Printing 10 Computer
design January- printer
12th
January
5 Secondary 13th January 13 20 8 Internet 250 Printer
data collection - Accessing stationery
20th
January
6 Primary data 21st 21 21 1 Printing 150 Computer

collection January excel,
printer
7 Sampling 22ndJanuar 22 30 9 Printing Computer
y- excel
30th
January
8 Data 1stFebruary 31 44 14 Field 600 Computer

c
c
c c
gathering - work excel

14thFebruar
y
9 Data 15thFebruar 45 49 5 Excel and
processing y- SPSS
19th
February
10 Data analysis 20thFebruar 50 58 9 Excel and
y- SPSS
28th
February
11 Conclusion 1st March- 59 68 10 Printing 40 Computer
10th March word
12 Report writing 11th March- 69 79 10 Printing 150 Computer
20th March word
13 Presentation 31st March 91 91 1

c
c
c c
c
cc
c
c
c
c
c
c
c
c c

mmm
1. c &)++: We had time constraints as being students; we were

involved in other assignments as well. Due to which we could not select a
larger sample size.This was the major constraint. Time for study on the project
is only 3 months
2. c 5$3")++: We also faced a budget constraint as this research is not

sponsored by any organization & is meant for academic purposes.
3. c )"7()++: We conducted the research only in Mumbai. This

also limited to administer the research to a larger survey.The results therefore
are confined to this area only and need not necessarily be applicable to other
areas.
4.c +)()" )++ The non-availability of the software for data analysis also
proved to be a limiting factor.
5. c 53 2 Respondents might have some biased views and this could
have affected the findings to a certain extent.
6. c )$7/The group members was only 4.


c
c
c c
mmm ,m > m

Ôc Laptops
Ôc Computers with Internet connection
Ôc Printers
Ôc Xerox machines
Ôc Pen Drives
Ôc Mobiles
Ôc Emails
c


c
c
c c
PART E
APPENDICES

c
c
c c
,, m

A.c 5
'( ) & ," )

)
1 1 OLS & Instrument Variable Panel Regressions 90
2 2 First Difference & Fixed Effect Panel Regressions 91
3 3 OLS Panel Regressions of Pathways 92
4 4 OLS Panel Regressions from Female Illiteracy 93
5 5 OLS Panel Regressions for Confounding Variables 94
6 6.1 Data Collection : Occupation 112

6.2 Data Collection : Child Weaning 112
6.3 Data Collection : Education Level 113
6.4 Data Collection : Income 113
6.5 Data Collection : Height 114
6.6 Data Collection : Weight 114
6.7 Data Collection : Frequency of Meals 115
6.8 Data Collection : Frequency of Health Check-Ups 115
6.9 Data Collection : Place of Treatment 116
7 7 PERT CHART 136

c
c
c c
5c 6
Sr. No. Chart No. Name Page No.
1 1.1 Casual Framework for Child Malnutrition 16
2 1.2 Proportional Mortality in Children 17
3 1.3 Child Mortality for selected diseases 21
4 1.4 Predicted Mean Ages 22
5 1.5 Deviation From Sex-Specific Mean Rates 22
6 1.6 Change in Child Stunting and Underweight 62
7 1.7 Child Underweight Survey across States 63
8 1.8 Child Stunting by Wealth Quintile 64
9 1.9 Child Stunting by Mothers Education 65
10 1.10 Correlation between Reduction in Poverty 78

& Growth


c
c
c c
> m m

c &

c +3
Ôc Male
Ôc Female
c "")$7
Ôc 15 - 18
Ôc 18 ² 26
Ôc 26 ² 41
Ôc 41 ² 55
Ôc Above 55
4.c ($
Ôc Single
Ôc Married
Ôc Divorced
Ôc Widower
Ôc Refuse to Answer
4c 6"%()*3$)+

Ôc Primary
Ôc Graduate
Ôc Post Graduate
Îc .)$&7() 39

Ôc Full Time
Ôc Part Time
Ôc Unemployed
åc !(%()* )$
&( m+)&9
Ôc < 50000
Ôc 50000 ² 100000
Ôc 100000 ² 300000
Ôc 300000 ² 500000
Ôc > 500000
c ! )$6"9
Ôc Below 4 feet
Ôc 4 ² 5 feet
Ôc 5 ² 6 feet
Ôc 6 feet & above

c
c
c c
c ! )$!"9

Ôc Below 30 kg
Ôc 30 ² 40 kg
Ôc 40 ² 50 kg
Ôc 50 & above
c.)$
Ôc Vegetarian
Ôc Non ² Vegetarian
c6)2)*+3) )$")*)6(1 =$79

Ôc Monthly
Ôc Quarterly
Ôc Yearly
Ôc Till u get Sick
c6)2&+ (3) )$%+ 9

Ôc 1
Ôc 2
Ôc 3
Ôc 4
Ôc 5

c6)2&+ &&'+ )$*&( 9

BBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBB

0c) )$+($3
$C(17)3$9
Ôc Yes
Ôc No
4c! )$&)+( #7+3$)+

))3m&9


Îc! )$)$7)+9


c
c
c c

åc3:$+)+*(+ )$()( 9
Ôc Yes
Ôc No
c )$7+&(+)$39
Ôc Yes
Ôc No
c!+33 )$(32++"9*7++")(3+ '()2")*å

@


c) )$2++ 2 )&1*)3+19
Ôc Yes
Ôc No
Ôc Don·t know
c!+&&')* )$
&( "123) )$"+(( ")*)
&+9
Ôc Public
Ôc Private
Ôc Others
c
+1 )$*) )$%($'(&DDD
c
c


c
c
c c
5m5m ,6.
1. Abler, D.G., G.S. Tolley and G.K. Kriplani, 1994: j

Westview Press: Boulder CO.
2. ACC/SCN (Administrative Committee on Coordination/Sub-Committee on
Nutrition), 2003: 5

. Draft.United Nations.
Nutrition), 2000a: º

. New York: United
Nations, in collaboration with the International Food
Policy Research Institute, Washington D.C.
Nutrition), 2000b,

º !" ###
$ %

&
& & , eds. J.Pojda, and L. Kelley, Geneva:
ACC/SCN in collaboration with ICDDR,B, Nutrition Policy Paper #18.
5. Adair, L., 1999: Filipino children exhibit catch-up growth from age 2 to 12 years,
"

%
,1140-1148.
6. Adato, M., R. Meinzen-Dick, L. Haddad, P. Hazell, 2003, Impacts of Agricultural
Research on PovertyReduction: Findings of an Integrated Economic and Social
Analysis, IFPRI: Washington, DC.
7. Aduayom, D., and L. Smith, 2003: Estimating undernourishment with household
expenditure surveys: Acomparison of methods using data from three sub-Saharan
African countries. In, ,

% %
-

, FAO, Rome.
8. Ahn, N. and A. Shariff, 1995: Determinants of Child Height in Uganda: A
Consideration of the SelectionBias Caused by Child Mortality, a

Î(1), 49-59.
9. Alaii, N., et al.., 2003: Perceptions of Bed Nets and Malaria Prevention Before and
After RandomizedControlled Trial of Permethrin-Treated Bed Nets in Western
Kenya, "

%j
, Î (Suppl), 142-48.
Philadelphia, mimeo.
10.Behrman, J. R., J. Hoddinott, J. A. Maluccio, Agnes Quisumbing, R. Martorell and
Aryeh D. Stein, The Impact of Experimental Nutritional Interventions on Education
into Adulthood in Rural Guatemala:Preliminary Longitudinal Analysis, 2003:
University of Pennsylvania, IFPRI, Emory Philadelphia-Washington-Atlanta,
processed.
11. Behrman, J. R. and J. C. Knowles, 1998b: The Distributional Implications of
Government FamilyPlanning and Reproductive Health Services in Vietnam,
prepared for the Rockefeller Foundation,University of Pennsylvania, Philadelphia,
mimeo.
c
c
c c
12. Behrman, J. R. and J. C. Knowles, 2003: Economic Evaluation of Investments in

Youth in Selected SEE Countries, Bangkok and Philadelphia, PA: University of
Pennsylvania (Report prepared for the World Bank Europe and Central Asia Region
² Social Development Initiative), processed.
13. Behrman, J. and M. Rosenzweig, 2004: Returns to Birthweight, .
%
)

(forthcoming).
14. Behrman, J. R., PiyaliSengupta and P. Todd, Progressing through PROGRESA:
An Impact Assessmentof Mexico·s School Subsidy Experiment, 2003: University of
Pennsylvania, Philadelphia, processed.
15. Bhargava, A., H. Bouis, and N. Schrimshaw, 2001: Dietary intakes and
socioeconomic factors areassociated with the hemoglobin concentration of
Bangladeshi women, "

%
: 758-764.
16. Bhutta, A., M. Cleves, P. Casey, M. Cradock and K. J. Anand, 2002: Cognitive and
Behavioral Outcomes of School-Aged Children Who Were Born Preterm, "

%

, (6):
17. Rosenzweig, M. R. and K. J. Wolpin, 1982: Governmental Interventions and
Household Behavior in aDeveloping Country: Anticipating the Unanticipated
Consequences of Social Programs, "

%& /
)

(2): 209-226.
18. Rosenzweig, M. R. and K. J. Wolpin, 1986, Evaluating the Effects of Optimally
Distributed PublicPrograms, )

0 åÎ*3): 470-487.
19. Rouse, D. 2003: Potential Cost-Effectiveness of Nutrition Interventions to Prevent
Adverse PregnancyOutcomes in the Developing World, "

%
,
:1640S-1644S.
20. Ruel, M. 2001: Can Food Based Strategies Help Reduce Vitamin A and Iron
Deficiencies? A Review ofRecent Evidence.Food Policy Review #5, International
Food Policy Research Institute, Washington DC.
21. Runge, C., B. Senauer, P. Pardey and M. Rosegrant, 2003: )

% a

3
Johns Hopkins University Press, Baltimore.
22. Sahn, D. and H. Alderman, 1988: The Effect of Human Capital on Wages, and the
Determinants of Labor Supply in a Developing Country, "

% & 1

)

*2): 157-184.
23. Saigal, S., L. Hoult, D. Streiner, and others, 2000: School Difficulties At
Adolescence In A RegionalCohort Of Children Who Were Extremely Low Birth
Weight, , , 4(2): 325-331.
24. Sastry, N. 1996: Community Characteristics, Individual and Hou sehold
Attributes, and Child SurvialinBrazil, &
(2): 211-229.
25. Sazawal S., R.E. Black, V.P. Menon, DinghraPratibha, L. Caulfield, U. Dhingra
and A. Bagati 2001:Zinc supplementation in infants born small for gestational age
reduces mortality: a prospective,randomized, controlled trial, , , :1280²6.

c
c
c c
26. Verhoeff, F.H., B.J. Brabin, S. van Buuren, L. Chimsuku, P. Kazembe, J.M.Wit,
R.L. Broadhead, 2001:An analysis of intra -uterine growth retardation in rural
Malawi, )
"

%$
,
27. Zimmerman, R. and M. Qaim. 2003: Potential health benefits of Golden Rice: A
Philippine case study. Mimeo. Bonn: Center for Development Research (ZEF),
University of Bonn.

c
c

RM Project Final 30.3

Transféré par

Informations du document

Description originale:

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

RM Project Final 30.3

Transféré par

Droits d'auteur :

Formats disponibles

c c

Research Report on:

N. L. Dalmia Institute of Management Studies and Research c

N. L. Dalmia Institute of Management Studies and Research c

As per your directions in Letter of Authorization, we have completed our research

N. L. Dalmia Institute of Management Studies and Research c

Research Report on:

N. L. Dalmia Institute of Management Studies and Research c

Subject: Letter of authorization to conduct research

N. L. Dalmia Institute of Management Studies and Research c

Also we would like to thank Prof. P. L. Arya, Director, N. L. Dalmia Institute of

N. L. Dalmia Institute of Management Studies and Research c

N. L. Dalmia Institute of Management Studies and Research c

6 ,-,, m 139

N. L. Dalmia Institute of Management Studies and Research c

N. L. Dalmia Institute of Management Studies and Research c

N. L. Dalmia Institute of Management Studies and Research c

N. L. Dalmia Institute of Management Studies and Research c

1. c H0 There is no significant effect of lower income levels on

2. c H0 There is no significant effect of inflation on malnutrition rate.

H1 There is significant effect of inflation on malnutrition rate.

3. c H0 There is no significant effect of poor eating habits on

4. c H0 There is no significant effect of inefficient Public Distribution

5. c H0 There is no significant effect of inadequate storage facilities on

N. L. Dalmia Institute of Management Studies and Research c

3. ACC/SCN (Administrative Committee on Coordination/Sub-Committee on Nutrition),

N. L. Dalmia Institute of Management Studies and Research c

6.Adato, M., R. Meinzen-Dick, L. Haddad, P. Hazell, 2003, Impacts of Agricultural Research

N. L. Dalmia Institute of Management Studies and Research c

N. L. Dalmia Institute of Management Studies and Research c

Although the etiology of IUGR is complex, a major determinant ofIUGR in developing

8. Ahn, N. and A. Shariff, 1995: Determinants of Child Height in Uganda: A Consideration of

N. L. Dalmia Institute of Management Studies and Research c

N. L. Dalmia Institute of Management Studies and Research c

To illustrate how to calculate the national burden of malnutrition, we give a step-by-step

N. L. Dalmia Institute of Management Studies and Research c

In all of these outcomes there is a doseresponse relationship with malnutrition. Specific

N. L. Dalmia Institute of Management Studies and Research c

potentiatedthe effects of infectious diseases on child mortality at population level

N. L. Dalmia Institute of Management Studies and Research c

N. L. Dalmia Institute of Management Studies and Research c

According to the results, underweight status among preschool-age children was

N. L. Dalmia Institute of Management Studies and Research c

More details on the quantitative relationshipsare provided in Section 3. A recent analysis

N. L. Dalmia Institute of Management Studies and Research c

Haddad, et al. (2003) estimate that the cross-country elasticity of pre-schoolunderweight

+(51")$+3)((+"+3 77)$+ )*

29. Barker, D.J.P., 1998:

N. L. Dalmia Institute of Management Studies and Research c

    , 107-118, The Johns Hopkins University Press,

Many poor nutritional outcomes begin   

In pre-school and school-age children, nutritional status is often assessed in terms of

N. L. Dalmia Institute of Management Studies and Research c

Controlled experiments with animals suggests that malnutrition results in irreversible

N. L. Dalmia Institute of Management Studies and Research c

41. Bourdillon, M., P. Hebinck, J. Hoddinott, B. Kinsey, J. Marondo, N. Mudege, and T.

N. L. Dalmia Institute of Management Studies and Research c

N. L. Dalmia Institute of Management Studies and Research c

N. L. Dalmia Institute of Management Studies and Research c

N. L. Dalmia Institute of Management Studies and Research c

N. L. Dalmia Institute of Management Studies and Research c

51. Coale, A. and E. Hoover, 1958: ,

N. L. Dalmia Institute of Management Studies and Research c

6 ,-,, m 139

In all of these outcomes there is a doseresponse relationship with malnutrition. Specific

+(51")$+3)((+"+3 77)$+ )*

, 107-118, The Johns Hopkins University Press,

Many poor nutritional outcomes begin

77)$+(3)6$+"+3(+$)+

66. FAO (United Nations Food and Agriculture Organization), 2003: j

69. Galloway, R. and J. McGuire, 1996:Women Need?

3$+" +*+ &)( The probability of infant mortality is estimated to be

73. Glewwe, P., 1999: j

2 $ $

77)$+ -3$+",%(+)m)++ +&+3m)3+

86. Evidence from the Philippines, %

stimates of the incidence of child malnutrition across households according to´wealth+