• TABLE OF CONTENTS
HIDE
 Front Cover
 Title Page
 Table of Contents
 List of Tables
 Foreword
 Acknowledgement
 Introduction
 Background of public distribut...
 Data and methods
 Sample characteristics
 Factors determining rationed rice...
 Rationed rice consumption...
 Rationed rice consumption and household...
 Discussion
 Conclusions
 Reference
 Author bio
 Notes
 Back Cover
 Reprint permission notice














Group Title: Research report - International Food Policy Research Institute ; no. 5
Title: Impact of subsidized rice on food consumption and nutrition in Kerala
CITATION THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00085348/00001
 Material Information
Title: Impact of subsidized rice on food consumption and nutrition in Kerala
Series Title: Research report - International Food Policy Research Institute
Physical Description: 48 p. (p. 47-48 blank) : ill. ; 26 cm.
Language: English
Creator: Kumar, Shubh K
Publisher: International Food Policy Research Institute
Place of Publication: Washington
Publication Date: 1979
 Subjects
Subject: Nutrition surveys -- India -- Kerala   ( lcsh )
Food relief -- India -- Kerala   ( lcsh )
Rice -- India -- Kerala   ( lcsh )
Alimentation -- Enquêtes -- Inde -- Kerala   ( rvm )
Aide alimentaire -- Inde -- Kerala   ( rvm )
Riz -- Inde -- Kerala   ( rvm )
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
Spatial Coverage: India
 Notes
Bibliography: Bibliography: p. 43-45.
Statement of Responsibility: Shubh K. Kumar.
Funding: Research report (International Food Policy Research Institute) ;
 Record Information
Bibliographic ID: UF00085348
Volume ID: VID00001
Source Institution: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: oclc - 05066581
lccn - 79105471
isbn - 0896290069

Table of Contents
    Front Cover
        Front Cover 1
        Front Cover 2
    Title Page
        Page 1
        Page 2
        Page 3
        Page 4
    Table of Contents
        Page 5
    List of Tables
        Page 6
    Foreword
        Page 7
    Acknowledgement
        Page 8
    Introduction
        Page 9
        Page 10
    Background of public distribution
        Page 11
        Page 12
    Data and methods
        Page 13
        Page 14
        Page 15
    Sample characteristics
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
    Factors determining rationed rice consumption
        Page 29
        Page 30
        Page 31
    Rationed rice consumption and nutrition
        Page 32
        Page 33
        Page 34
        Page 35
    Rationed rice consumption and household caloric intake
        Page 36
        Page 37
        Page 38
    Discussion
        Page 39
        Page 40
        Page 41
    Conclusions
        Page 42
    Reference
        Page 43
        Page 44
        Page 45
    Author bio
        Page 46
    Notes
        Page 47
        Page 48
    Back Cover
        Page 49
        Page 50
    Reprint permission notice
        Page 51
Full Text









Impact .. u i idized Rice






b. byShubhi CK Kumar;






P. : 'i ; .. -
..c
.. : : .. *.











. t ,. Si.

!. ... :-,.
; : 7 : ...-A ,
'..p. I ". .



























































The International Food Policy Research Institute is an independent. nonprofit organization
that conducts research on policy problems related to the food needs of the developing world.
IFPRI's research is directed toward policy makers at the national and international level and
is distributed to those cocerned with food policy issues.

























IMPACT OF SUBSIDIZED RICE ON

FOOD CONSUMPTION AND NUTRITION IN KERALA




























































The research presented in IFPRI's Research Reports is conducted at the Institute; however, the
interpretations and views expressed are those of the authors and are not necessarily endorsed
by the Institute or the organizations that support its research.




















IMPACT OF SUBSIDIZED RICE ON
FOOD CONSUMPTION AND NUTRITION IN KERALA





Shubh K. Kumar





















Research Report 5
International Food Policy Research Institute
January 1979





































































Copyright 1979 International Food Policy Research Institute.

All rights reserved. Sections of this report may be reproduced with
acknowledgement to the International Food Policy Research Institute.




ISBN 0-89629-006-9













CONTENTS




Foreword
1. Introduction 9
2. Background of Public Distribution 11
3. Data and Methods 13
4. Sample Characteristics 16
5. Factors Determining Rationed Rice Consumption 29
6. Rationed Rice Consumption and Nutrition 32
7. Rationed Rice Consumption and Household Caloric Intake 36
8. Discussion 39
9. Conclusions and Implications for Research 42
References 43













TABLES



1. Observation period 15
2. Comparison of income distribution in the
sampled households with rural Kerala 17
3. Size distribution of holdings in Trivandrum
District and in sampled households 17
4. Occupational distribution of workers 18
5. Caloric and protein composition of house-
hold diets 20
6. Seasonal distribution of household food
expenditures 23
7. Overall dietary measures by income groups
and subgroups 24
8. Spatial differences in ration rice consumption 30
9. Nutritional impact of household incomes and
ration subsidy 34
10. Income and ration subsidy effects on house-
hold caloric intakes 37








FOREWORD


It is increasingly clear that enlarged
food supplies are not automatically ac-
companied by commensurate increases in
effective demand for food and that this
deficiency in demand can be traced to
inadequate growth in purchasing power
of low income people, who spend the
bulk of increments to income on food. A
match between growth in food produc-
tion and effective demand for food is
necessary not only to meet humanitarian
objectives, but to maintain the remuner-
ative prices essential to rapid growth in
food production as well.
The International Food Policy Research
Institute is initiating a major research
program to analyze the linkages among
growth in agricultural production and em-
ployment in ancillary producer and con-
sumer goods industries, income growth of
low income people, and demand for food.
Of particular importance in this research
is the analysis of expenditure patterns,
particularly the expenditure patterns of
low income families, who spend a high
proportion of their income increments on
food. This research is expected to suggest
policies to strengthen the complex growth
fostering linkages between food produc-
tion and consumption.
The food ration-subsidy schemes em-
ployed on a large scale in South Asia
are important efforts to directly increase
effective demand for food. They have par-
ticular relevance to such issues as adapting
food consumption to increased productiv-
ity growth rates; to increased availability
of food aid; and to policies for raising
the incomes, food consumption, and wel-
fare of low income families. The Inter-
national Food Policy Research Institute is
engaged in a comprehensive analysis of


variants of such programs in Bangladesh,
India, and Sri Lanka to determine their
impact on consumption and welfare in the
region and to learn lessons for application
in other areas.
This study by Shubh Kumar is part of
that larger effort. It is based on detailed
study of a small sample of low income
families in Kerala, India. It includes data
on access to rationed, subsidized food;
food consumption; and nutritional status
of infants as measured by height and
weight. Establishing and measuring the re-
lationship among these variables under
field conditions is a seminal contribution
which points the way to more extensive
future work along similar lines. More
broadly, Shubh Kumar's work provides
data on the various effects of income from
different sources on food expenditure and
health. This knowledge is of considerable
value for designing programs to improve
food consumption and is of particular rel-
evance if reliable, large scale food aid
programs are to be instituted with the ob-
jective of improving the welfare of low
income families. Shubh Kumar's profes-
sional background in nutrition, her more
recent work in economics, and the intui-
tive insight that comes from her close
working relationship with the families in
her sample, combine to make this an im-
portant work in an area of special interest.
Shubh Kumar plans to expand this research
to cover different types of situations and
relate it to a broader range of health and
welfare conditions as part of a set of IFPRI
studies in this area.


John W. Mellor


Washington, D.C.
December, 1978

























ACKNOWLEDGEMENTS
I wish to acknowledge the support of John Mellor, who helped set the course of this
study and provided encouragement throughout, and Michael Latham, of Cornell Univer-
sity, who participated in the earlier stages of this study.
My thanks to Jim Gavan for his interest and helpful suggestions in the later phases
of the study. I am also grateful for the useful comments of Shlomo Reutlinger, Paul
Isenman, Dharm Narain, Barbara Huddleston, Emma Simmons, and Marcelo Selowsky,
who reviewed this study.
In addition, thanks goto Phil Willems and Kate Hathaway for their editorial comments,
and to Ruth Ellen Rounds for typing the manuscript.










1




INTRODUCTION


This paper empirically analyzes the im-
pact of a food price subsidy program on
levels of food consumption and nutrition
of a low income population.
Specifically, three related questions
are investigated:
1) What are the factors that determine
households' access to the subsidized
food-in this case rationed rice (8,
34);
2) What is the impact of the subsidized
food on levels of household nutri-
tional intake, child nutritional status,
and consumption of major staples;
3) Howdoestheincomesubsidy implied
bythe subsidized food compare with
increments in other sources of in-
come.
As the above statement of objectives
implies, the scope of this paper is quite
narrow. By focusing mainly on nutritional
benefits, the analysis limits its conclu-
sions to the welfare aspect of subsidized
food distribution. To a lesser extent it also
considers the association of subsidized.
food with changes in diet composition
and in food demand. It is recognized,
though not considered in this paper, that
linkages of such a program with other
sectors of the economy and the overall
costs involved are essential for making
informed decisions on any food policy or
program. To give decisionmakers a com-
plete perspective, however, some mea-
sure of direct welfare implications is
equally necessary.


Nearly every country in the world today
recognizes adequate nutrition for its
people as a goal. The most pervasive form
of nutritional deficiency is energy pro-
tein malnutrition (32). Subsidized food
distribution programs areof special interest
in addressing that problem because they
have the potential to increase real in-
comes and food consumption of low
income groups while at the same time in-
creasing aggregate demand for food. Even
without the additional cost of explicit
targeting, such programs sometimes
achieve that effect (29). And while in-
creasing incomes by itself may not be the
solution, for low income groups it is the
constraintof inadequate purchasing power
that limits the consumption of food and
is responsible for many of the nutritional
problems in the developing world.
The present analysis of food consump-
tion and nutritional status over a six
month period uses observations from a
small sample of low income households
in India's Kerala State compiled by the
author between February and September
1974. This work in part continues the au-
thor's previous work on the components
of household income that determine child
nutrition among low income rural house-
holds. Previous analysis indicated that
aggregate household incomes had prac-
tically no association with child nutri-
tional status, but that disaggregating in-
come by source showed that increasing
certain income components could sub-









stantially improve child nutrition (17).
The present analysis extends that work in
two ways. First, it includes the income
transfer implicit in the consumption of
subsidized, rationed rice as an income
source-to compare its effects on child


nutrition with those of income from other
sources. Second, it expands the study by
including the effect of added income on
the adequacy and composition of house-
hold diet as well as on child nutrition.










2



BACKGROUND OF PUBLIC

DISTRIBUTION


As a means of achieving distributive
justice, universal public food distribution
has found a new respectability. In two
areas where it has been widely used-in
Kerala and in Sri Lanka-a quality of life
above that of other areas with compar-
able income levels has been attained.
Even though these two cases have other
common characteristics, e.g., a high
availability of health and education fa-
cilities that doubtless contribute to their
high rating on the Physical Quality of Life
Index (PQLI)*, dietary quality by itself is
also likely to be a factor.
Such measures of life quality as infant
mortality rates have only an indirect, ten-
uous link with improved food availability
at the aggregate level. However, the link
between food availability and, for ex-
ample, infant mortality rates does exist-
via levels of maternal nutrition (14). Mea-
suring infant mortality that results from
maternal malnutrition, namely premature
births and low-weight-for-gestational-age
births, would provide a direct causal link
with the level of food availability.t It is
possible, therefore, to correlate access


to food from the public distribution sys-
tems and health by studying the subsidy's
immediate link with food consumption
and especially with nutritional status.
Given the wide variation among popu-
lations of countries in achievement levels
for food consumption relative to needs,
health, and education, it is obvious that
if there is a wider availability for all
groups, it would be reflected in the aggre-
gate achievement levels. By directing sup-
plies to those with the highest income
elasticities of demand, public food dis-
tribution systems may have the potential
of distributing food to meet human re-
quirements better, while also increasing
aggregate food demand. The strongest
arguments against food subsidies are
made in terms of the production disin-
centives created by supplying food at
lower prices, and the very visible fiscal
cost involved.
In India, a combination of levy pro-
curement and price support, often with
restrictions on regional grain movements
during country-wide production short-
falls, has been used to supply the public


PQLI, developed by the Overseas Development Council, is derived from the average rating achieved by a country
for the three indicators of life quality-life expectancy, infant mortality, and literacy rates. Lower income countries with
an average per capital GNP of $152 in the mid-1970s had an average PQLI of 39. Kerala had $110 per capital income
and a PQLI of 69; Sri Lanka, $130 per capital income and a PQLI of 83 (24).
t A more direct connection has also been made between increased availability of formula foods for infants and
higher infant mortality rates (25).









food distribution system.* However, both
"a priori reasoning and facts of the Indian
situation militate against the view that
the burden of protecting the poor in a
dual market system falls on the pro-
ducer." (2) Although the importance of
prices is well established (21, 22) still,
many unresolved questions remain on the
costs, distribution of benefits, and incen-
tives for production resulting from pric-
ing policies and operations needed for
public food stocks and food distribution.
For example, there has been practically
no investigation into the effect of such
policies on production and marketing of
new grain varieties. Moreover, such new
varieties do not always fit readily into
existing consumer preference patterns,
and procurement and subsidized distri-
bution may have played a crucial role in
some areas in expanding markets for con-
sumption of the new varieties both among
lower income groups habitually consum-
ing that foodgrain and among populations


not traditionally consuming it.
No studies now available attempt to
empirically evaluate the impact of sub-
sidized food distribution on consumption
and nutrition for different income groups
(7, 8). Three factors are expected to de-
termine the actual impact on consump-
tion. First, the place of the commodity
in the diets of different income groups is
important; those most likely to consume
it will benefit most. Second, explicit or
implicit (by commodity selection) at-
tempts to target the distribution of food
will affect consumption. In India, public
food distribution has had the stated ob-
jective of reaching only urban and food
deficit areas. Targeting may also try to
limit consumption of the subsidized food
to certain income groups, as Sri Lanka is
currently attempting to do. Finally, the
propensity to consume additional food
from the food subsidy income determines
the real improvement in consumption and
nutrition.


Since in the Indian case, procurement prices have consistently been set to cover production costs and margins,
with zoning in past poor years and without it in the present context, state purchases in surplus areas can be seen as a
price support activity reducing open market supplies and pushing up the price. In the present context with a comfortable
food supply situation, and the elimination of zonal restrictions, the procurement operations are more clearly a price
support operation (12). Thus even though the government has been shown to purchase grain at somewhat below
average open-market prices in India, it has also been shown that the weighted average of prices received by farmers
is higher in the presence of the Government's operation even in a surplus rice region (10, 31). At the same time the
supply situation in food deficit regions such as Kerala improved with the transfers on government account over and
above that expected by purely open market mechanisms (8).










3





DATA AND METHODS


A survey conducted by the author in
1974 in rural Trivandrum District to cover
the breadth of the midland agricultural
economy provides the data source in the
present analysis. A baseline survey was
conducted on a stratified random sample
of 120 households from three sites in the
district (Figure 1). For the intensive sur-
vey, the baseline sample was narrowed to
43 selected households that:
1) represented the bottom 50-60 per-
cent of income levels, and
2) had children of weaning age, i.e.,
between six and thirty-six months.
The survey was conducted between
February and September 1974, and for
six months a continuous record was kept
of income and employment by house-
hold members; own-farm income; food
expenditures and purchased quantities of
certain items; food intake; morbidity pat-
terns; and nutritional status of weaning-
age children in each household. The tim-
ing of observations made in the study is
shown in Table 1. The analysis in this
paper draws only on data categories 1,
12, 13, and 15, and is restricted to the in-
tensive survey sample. When the survey
was conducted, its objective was to as-
semble information on the entire system
of effects that determine child nutrition
in a low income environment. The survey


was not designed explicitly to study the
subsidized food distribution system. The
distinction between ration and open mar-
ket rice emerged when interviewers re-
corded observations in the weekly food
purchase record of every consumed item,
detailing quantity, price, frequency of
purchase, source of purchase, and terms
(cash, barter or credit) of purchase for
each commodity. Because of the care
taken in the recording and storage of
household interviews in the original form,
the data base was available for this
analysis.
Such a small sample of households has
some disadvantages. Among the foremost
is the limitation it places on statistical
interpretation of any multivariate analysis,
so that only simple hypotheses and re-
lationships can be tested. It also reduces
the ability to generalize to larger popula-
tion groups. The advantage, however, is
that measurements are more accurate
and a much wider array of characteristics
can be recorded. This permits a much
fuller understanding of the sample and
leads to greater sensitivity not only in
interpretation of results, but also in the
formulation of hypotheses to be tested
for larger surveys. Given the high cost of
nationwide research, the value of smaller
studies in the formulation of an adequate
research design should be apparent.





Figure 1 TRIVANDRUM DISTRICT AND LOCATION OF STUDY AREAS
Location of Study Areas
( Koonthaloor and Puravoor
Vellanad
SThollcode












Nedumangad .

Arabian Sea




Kerala T

SN6yyattin ara/





District Boundary India
Roads
*."*#* Railway
-..-4 River
O Capital
OTaluk Headquarters & Town
Trivandr Miles 1
5
Kilometres IiUi .' l









Table 1 -Observation period

Feb.-Mar. April Mar.-Jun. May-Jun. Jun.-Sep. Sep. Sep.
Data Categories Schedule 1' NS 1b Schedule IIC Schedule Schedule II' Schedule NS2b
(Mi-Ms)d IlI (M6-M,)d Ill'


I. Child
1. Anthropometric measurements X X
2. Clinical examination X X
3. Dietary intake X X
4. Disease incidence X X
5. Weaning and disease history X
II. Mother
6. Activity time inventory X X
7. Exposure to media, services X X
8. Employment X X
III. Family
9. Social, demographic X
10. Housing, health, sanitation X
11. Land, cultivation X
12. Farm output X X X
13. Wage income and employment X X X
14. Consumption expenditure, major items X X X
15. Detailed food purchase record X X
16. Time allocation by member X X


SSchedule I: baseline survey.
SNS 1, NS 2: assessment of nutritional status.
Schedules II, Ill: intensive survey of low income households.
M1 -Mg : periodic revisits.
U-1










4





SAMPLE CHARACTERISTICS



Levels of Living

The major distinguishing characteristic
of the sample households was their low
incomes. Their average monthly income
was Rs. 33 per capital, while the Kerala
State monthly expenditure averaged Rs.
55 per capital. Table 2 shows the percent-
age distribution of the population by in-
come levels in the study sample and in
Kerala.* The lack of access to land for
cultivation also reflects the low economic
status of the studied households. All were
either landless or near-landless; with all
but one household depending on off-
farm income sources. The size distribu-
tion of access to land for cultivation in
the study sample and in Trivandrum Dis-
trict is shown in Table 3.t
The occupation of workers in the sample
households is shown in Table 4. Of the
312 household members in the sample,
37 percent participated in the labor force;
the corresponding figure is 29 percent
(1971) for Kerala State (34, p. 74). The
higher participation rate, which is to be
expected at lower income levels, was in-
terpreted as a result of a higher propor-
tion of women in the wage labor force-
about 30 percent for the sample versus


13.5 percent for Kerala State. Table 4 also
reveals the overwhelming predominance
of casual and migrant workers in the labor
force. Working for wages was the primary
source of livelihood for all but a small
fraction of the sample. Farming was a
major occupation for only five males (4.3
percent of the labor force) and for none
of the females. This is expected given the
small size of land holdings.

Composition and Adequacy of
Household Diets
When consumption of the sample
households is averaged for the six month
period between March and September
1974, rice is found to contribute about 34
percent of the calories and 33 percent of
the proteins in the diet, while tapioca
contributes nearly 41 percent of dietary
calories and 13 percent of dietary pro-
teins (Table 5). Ration rice makes up 60
percent of rice consumed by these house-
holds, compared with about 33 percent
for the state (8). In percentage terms,
therefore, the contribution of ration rice
to food consumption for these lower in-
come households is much more important
than it is for the state as a whole. Since
the coverage of the ration systemS in


Since income measures in the study sample are averaged over income flows during a nine-month period, they are
more likely to be comparable with expenditure data, for low income groups.
t Trivandrum District is second only to Allepy district in Kerala State in population density and predominance of
small-sized holdings.
At least in terms of number of households covered by each ration outlet.










Table 2-Comparison of income distribution in the sampled households
with rural Kerala


Rural Kerala Study Sample
(Expenditures) (Incomes)b
Rupees per Capita
per Month
Percentage of Cumulative Percentage of Cumulative
Households Percentage Households Percentage


less than 15 1.2 1.2 18.6 18.6
15-24 8.4 9.6 16.3 34.9
25-34 13.8 23.4 32.5 67.4
35-74 51.6 75.0 28.0 95.4
more than 75 25.0 100.0 4.6 100.0

100.0 100.0


'National Sample Survey, Twenty-eighth Round, October 1973-june 1974.
bDecember 1973-September 1974. Based on household averages of nine monthly income measurements.









Table 3-Size distribution of holdings in Trivandrum District and in
sampled households


Trivandrum District Study Sample
Acres Percent Holdings Percent


less than
0.10 nab 16 37.5
.11- .60 55.5 22 52.1
.61-1.30 22.3 4 8.3
1.31-2.60 13.7 1 2.1
2.61-5.25 6.3 0 0
5.26-7.90 1.4 0 0
more than
7.90 0.8 0 0

100 43 100


Source: Kerala State Bureau of Economics and Statistics, 1973.
1 acre = .404 hectare.
"The report enumerates only holdings .04 hectares and above and does not provide any estimate of landlessness.









c Table 4-Occupational distribution of workers

Employment of Males Employment of Females

Average No. with Nature of Average Average No. with Nature of Average
Primary Occupation No. Education Secondary Secondary Education No. Education Secondary Secondary Education
(years) Occupation Occupation (years) (years) Occupation Occupation (years)


1. Fixed wage


2. Farming


3 9.0


5 7.0


1 Casual agr.
labor


3 Casual agr.
labor (2)
Casual
skilled (1)


3. Self-employed in home
without capital
with capital



4. Self-employed outside home
without capital


1 3.0
6 7.6


4 4.3


0
1 Self-employed 8.0
in home with-
out capital


4 1.0
7 5.0


1 Casual nonagr. 5.0 10 1.4
labor


1 Casual agr. 0.0
labor


6 4.5 1 Farming 5.0 2 6.5


3 7.0


with capital











5. Casual agricultural labor


6. Casual worker-household
industry
7. Casual skilled worker
8. Casual nonagricultural
laborer
9. Migrant Workers
Fixed wage
Self-employed
Laborer
Total


19 3.4 2 Farming(1)




Casual nonagr.
labor (1)


4 6.3 0


4.0 15 1.9 7 Self-employed 0.4
outside home
without capital
(6)
Nonagr.
labor (1)


8 4.1 0


2 0.6 0


5.4 51 3.0 9










Table 5-Caloric and protein composition of household diets
(Average for March-September 1974)

Calories Grams of Proteins
Foods
Per Adult Percentage Per Adult Percentage
Equivalent of Equivalent of
Per Day Total Per Day Total


Cereals:
Rice
Ration 413 20.3 8.3 20.0
Open 276 13.6 5.4 13.0
Wheat 31 1.5 1.0 2.4
Subtotal 720 35.4 14.7 35.4


Tapioca 827 40.7 5.5 13.3
Grains and Legumes 29 1.4 2.0 4.8
Fruits and Vegetables 41 2.0 1.3 3.1
Milk 33 1.6 1.4 3.4
Fish, meat, eggs 54 2.7 10.1 24.3
Oils, oilseeds, sugar 128 6.3 0.0 0.0


Purchased meals 5 2.5 1.0 2.4
Nonpurchased meals artwork and feeding program 145 7.4 5.5 13.3
2,032 100 41.5 100.0


Recommended allowance for reference adult
males (adult equivalent) 2,530' 46"


SF.A.O. average requirement for a 'moderately active', adult male weighing 55 kg. -the suggested weight for the Indian
population according to the Indian Council of Medical Research (6, 12)
SThe safe level of intake for reference adults in a population is the upper limit of individual requirements, or the
amount of protein, "considered adequate to meet the physiological needs and maintain the health of nearly all
healthy individuals in the specified category" (p. 85). For the present diets, the estimated NPU = 72 is used to
derive the safe level of intake.








Trivandrum District where the study was
conducted is nearly the same as in the rest
of the state, a similar contribution by
ration rice could be expected for other
lower income households in Kerala. In
absolute terms, the study households ob-
tained an average of 99 grams per capital
per day of ration rice over the six months,
only slightly more than the average ration
of 92 grams per capital per day for Kerala
during that year (8).
Tapioca predominates as an inexpen-
sive calorie source for the households. It
is an inferior commodity, whose con-
sumption declines rapidly with increasing
income. For this sample, tapioca provided
about 40 percent of the caloric intake; for
the state as a whole, it provided about 25
percent (34). Even within this small study
area (see Figure 1) some intraregional
variation exists in many aspects of dietary
composition, including rationed rice con-
sumption, as described in a later section
of this paper. With 54 percent of its acre-
age in holdings of less than two and a
half acres, tapioca is an important sub-
sistence crop. Even though a substantial
proportion of the crop reaches the mar-
ket,* lower income households largely
consume it at home. For such households,
however, as much as half an acre of tap-
ioca can meet no more than 30 to 40 days
of their need.


In the present diets of these low income
households, the caloric deficit is about 20
percent, which is more than the primary
or apparent protein deficit of about 10
percent. At these levels of caloric intake,
however, secondary protein deficiency
would be greater than 10 percent, since
some of the proteins would be broken
down to meet the more basic energy re-
quirements. Even though within more
homogenous socioeconomic groups some
of the rationale for recommending a
higher, 'safe' level of intaket for proteins
is lost, the principle of predicting in-
equalities at low overall intake levels may
be applied within households in the lower
income groups, as it is between house-
holds in larger population groups. The
evidence therefore points towards a lower
than recommended protein and calorie
intake at the household level.
The high fish content in the Kerala diet
is responsible for better overall protein
quality in the diets of these low income
households than in the aggregate Indian
diet.t The proportions of proteins from
different sources based on individual
commodity information for these house-
holds for the six months were:
Cereals 42 percent
Animal sources 33 percent
Other plant sources 25 percent
Efficiency of Nitrogen Utilization (NPU)
72**


Especially since those who produce more will have higher incomes and are likely to consume less than the
smaller producers of the crop, and therefore sell the major proportion of their crop.
t A "safe" level of intake is calculated by adding twice the coefficient of variation to the average requirement of
proteins, some vitamins, and minerals.
$ Fish was a very important component of the diet. Its absence, a rarity, occurred because either eggs or meat were
consumed that day or because of extreme impoverishment. In the interior midland sites, though fish was trucked in
once daily from the coast, dried fish was often used by the lower income households especially during the riionsoons
when fresh fish became relatively scarce.
**Net Protein Utilization (NPU) is a combined measure of digestibility and of the efficiency of utilization of the
absorbed amino acids. It is derived from nitrogen balance studies in humans under different dietary regimes. Even
though it is considered inferior to the amino acid scoring pattern for determining the protein quality of a diet, it is a
satisfactory approximation especially when quantities for each food item consumed are not individually analyzed
(5, pp. 49, 70).








This can be compared with the protein
composition of the average Indian diet
(9):
Cereals 70 percent
Animal sources 9 percent
Other plant sources 21 percent
Efficiency of Nitrogen Utilization (NPU) =
61 (approx.)
The favorable combination of protein
sources in the diet implies that relatively
small increases in cereal consumption,
while narrowing the caloric gap, will even
more rapidly reduce the primary protein
gap as cereals replace tapioca in these
diets. An increase in cereal consumption
does not necessarily substitute some of
the tapioca calories in the diet, to the ex-
tent that the additional cereal would pro-
vide a more concentrated source of calo-
ries and proteins, and be fed to weaning-
age and young children, (while tapioca, if
possible, is avoided).* It would, in addi-
tion to reducing the calorie gap for the
households, tend to improve both quan-
tity and quality of proteins for the young,
for whom excessive bulk in food can limit
total nutritional intake. In protein quality
alone, rice is far superior to tapioca.t
Furthermore, rice provides five times as
much protein per unit of calorie ingested,
an important ratio from the standpoint of
child feeding and nutrition. Rice protein
also has one of the highest amino acid
scores among cereals, and because of its
easy digestibility forms a traditionally
preferred base forweaning foods in Kerala.


As in most developing countries, children
in rural India are often weaned almost di-
rectly to the adult diet (37), especially at
the lower end of the income scale, where
the simplicity of the diet makes the nu-
tritional contribution of the major staple
even more important.

Seasonal Variations in Food
Consumption

The survey for which results are being
reported covered the months of March-
April to August-September. Agriculture
provides employment for only 62 percent
of the rural labor force in Kerala (com-
pared with 85 percent for India).[ For the
unskilled casual rural workers who form
the bulk of the landless and near landless,
it is probably the most remunerative em-
ployment source. Much of the employ-
ment that these unskilled workers seek in
other activities is therefore determined by
the rhythms of the agricultural cycles.
The present survey began during a pre-
rice-planting slack season (March-April).
Agricultural activity was brisk in April-
May and May-June with field preparation
and transplanting underway. In June-
July activity in the rice fields began taper-
ing off, with mainly weeding occurring
as tapioca harvesting and related employ-
ment picked up. July-August was a gen-
erally slack period when practically no
field work was done. August-September
was the rice harvesting period.


Kerala has an accurate local vocabulary for the Kwashiorkor syndrome which is called "Neeru Karupan." "Neeru"
is derived from Sanskrit, meaning water, and "Karupan" denotes lesions and flakiness of the skin. Kwashiorkor is
associated with oedema and dermatosis. The food belief prevalent among mothers in Kerala, which is doubtless the
result of observation, is that if children are fed liberally on tapioca, they develop "Neeru Karupan" (26, p. 615). In
the present study, it was observed that tapioca was not given in the early weaning period and was usually introduced
gradually after about two years of age.
t The NPU value for rice is about 65, it is about 50 for whole wheat. Protein quality, based on the amino acid
score, is 67 for rice, 53 for whole wheat and about 35 for tapioca (5, 12).
t According to the 1971 Census.








The seasonal distribution of total house-
hold food expenditures does to some ex-
tent parallel the swings of the rural, agri-
culture-based economy (Table 6). Thus
the smaller food budgets in March-April
and July-August reflect lower agricultural
employment and incomes at that time.
The higher values of food consumption
between April-May and June-July are the
combined result of higher incomes from
wages during the earlier months and from
the availability of home produced food
during the later part of this period. For
low income rural households in Kerala
that operate small plots, usually of dry
land suitable only for tapioca and horti-
culture, the primary source of income is
wages. Furthermore, the harvest of much


of their farm produce (for example, tap-
ioca) coincides with the period of slack
wage employment and income. This is in
marked contrast to the farmer-centered
view of the rural economy, which sees
wage income as an off-season supplement
to the more fundamental farm income
source.



Income and Intraregional
Variations in Food Consumption

Indicators of dietary composition and
nutrition by income levels, food expendi-
ture categories and subregional location
of households are shown in Table 7. The
main observations are:


Table 6-Seasonal distribution of household food expenditures

Categories of March- April- May- June- July- August-
Food Expenditure April May June July August September

(Rs. Per Household Per Month)
1. Total food expenditure 30.8 37.5 36.1 37.1 34.7 63.1"
2. Ration rice 5.0 5.9 5.6 6.5 7.0 7.7
3. Open market rice 8.6 10.7 9.1 8.7 5.2 16.8
4. Tapioca 6.9 6.1 4.4 4.8 7.2 5.6
5. Vegetables and fruits 1.8 2.8 4.3 4.1 1.4 8.1
6. Milk and fresh foods 3.8 5.6 4.9 6.3 6.3 6.7
7. Condiments 2.7 3.9 3.9 4.2 4.6 13.0
8. Purchased meals 1.4 1.6 3.2 1.6 1.7 0.8
9. Value of home produced consumption
(excluding rice)' 1.52 2.76 6.36 6.32 4.64 5.13


a This figure was pushed up by large "Onam" purchases by some households. "Onam" is a festival, characterized by
feasts, that falls immediately after the harvest in Kerala (usually between late August and September).
'' Additive of home produce value in items 4 through 7.










Table 7-Overall dietary measures by income groups and subgroups


Total Sample
(N = 43)
Monthly Per Capita Total Income
(Rupees)
Less than 15-24 25-34 35-49 50-74 75 or
15 more


Food expenditure
Rs. per capital per month 15.3 20.2 26.8 28.4 29.7 29.1
Total calories (daily adult equivalent) 1,086 1,893 2,102 2,211 2,290 1,898
Total grams of protein
(daily adult equivalent) 29.9 36.4 43.9 49.4 50.6 47.0
Ration rice calories
(daily adult equivalent) 263 376 414 439 569 584
Open market rice calories
(daily adult equivalent) 133 240 336 338 270 386
Tapioca calories
(daily adult equivalent) 1,013 898 819 817 729 213
Index of diet quality 1.7 2.2 3.2 3.4 4.1 6.0


N-number of households


8 7 14 6 6 2


Subregion 1
(N = 24)
Monthly Per Capita Total Income
(Rupees)
Less than 15-24 25-34 35-49 50-74 75 or
15 more


Food expenditure
Rs. per capital per month 20.5 27.46 27.6 34.5 28.0 29.1
Total calories (daily adult equivalent) 1,784 1,933 1,934 2,374 2,224 1,898
Total grams of protein
(daily adult equivalent) 37.3 44.4 45.2 52.7 50.0 47.0
Ration rice calories
(daily adult equivalent) 522 544 474 551 590 584
Open market rice calories
(daily adult equivalent) 209 460 359 521 324 386
Tapioca calories
(daily adult equivalent) 655 440 518 630 692 213
Index of diet quality 2.7 3.1 3.2 3.2 4.1 6.0


1 3 10 3 5 2


N-number of households









Subregions 2 and 3
(N = 19)
Monthly Per Capita Total Income
(Rupees)
Less than 15-24 25-34 35-49 50-74 75 or
15 more

Food expenditure
Rs. per capital per month 14.5 14.6 24.9 22.4 37.9 -
Total calories (daily adult equivalent) 1,597 1,780 2,334 1,937 2,466 -
Total grams of protein
(daily adult equivalent) 26.5 30.4 40.5 46.2 53.6 -
Ration rice calories
(daily adult equivalent) 226 249 414 328 467 -
Open market rice calories
(daily adult equivalent) 123 75 277 155 0 -
Tapioca calories
(daily adult equivalent) 1,065 1,241 1,571 1,005 914 -
Index of diet quality 1.6 1.4 3.2 3.7 3.8 -
N-number of households 7 4 4 3 1 0


1) In most cases indicators of nutri-
tional adequacy for the household
diet, namely caloric and protein in-
takes, and the index of diet quality*
increase with increments of per
capital household income.
2) Consumption of rationed rice calo-
ries increases with cross-regional
increases in incomes and food ex-
penditures, but whether this occurs
within subregions is not clear.
3) Subregions 2 and 3 have a higher
incidence of the lowest income
households and also tend to have
lower overall consumption or ra-
tioned rice. Factors that determine


the consumption of rationed rice
will be addressed in greater detail
later.


Child Nutritional Status

For the young child, the single anthro-
pometric measure that best reflects over-
all growth and nutritional status is the
ratio of actual to expected weight for age.
In this study, it was possible to use this
indicator of child nutrition since the major
drawbackto its use, in accurate age assess-
ment, was overcome. The revised inquiry
format of this survey permitted cross-
checking reported age with date and


This index, derived for the present study, is based only on foods other than main caloric sources, and weights the
quantity consumed (measured by expenditure when foods were not purchased on a weight basis) over the week by their
frequency. This is based on the assumption that even a small quantity of these foods, with a staggered consumption
would provide greater improvements to diet quality, both in terms of protein quality of major calorie sources and
availability of vitamins and minerals for body functions, than if consumption were concentrated in a limited period.








month of birth on the local calendar.*
As mentioned earlier, all households in
the survey had at least one child of wean-
ing-age, six to thirty-six months. The nu-
tritional status of one child in this age
group in each household was recorded
twice during the study-once soon after
beginning and again five or six months
later at the end of the survey.
The distribution of initial weights and
heights relative to age of the children is
shown in Figures 2 and 3, together with
means for American and Indian children.
All the children in the sample are below
the Harvard 50th percentile weight stan-


dard and about 60 percent of them are
below the mean weight curve for Indian
children.t However, nearly 70 percent
are above the mean height curve for In-
dian children. This suggests that while
current nutritional status for these low in-
come children falls a little below the In-
dian average, their prenatal and neonatal
nutrition appear to have been superior,
giving them a better start on height.
The analysis in this paper uses the mean
of the two "weight for age" measures of
each child as an indicator of the child's
nutritional status.


Even for these children aged six to thirty-six months whose nutritional status was measured in this study, a
surprisingly high rate of age underreporting was found.
t These means for Indian children do not represent local standards and were simply measurements of apparently
healthy children without any clinical signs of nutrition deficiency (13). Since local standards for Kerala are not available.
and evidence suggests that well-nourished Indian children closely follow Harvard standards until at least five years,
these have been used in assessing nutritional status in this study.

















Figure 2 DISTRIBUTION OF CHILD WEIGHT FOR AGE


Harvard 50th percentile
(Jelliffe, 1966)


* **


** *


India-mean weights
(ICMR, 1972)


* *** *


* *


* *
(2)


**


* *


3 7 11 15 19 23 27 31 35 39 43

Age ( months) Rural Kerala
















Figure 3 DISTRIBUTION OF CHILD HEIGHT FOR AGE

96-


92-
India: mean heights
*^ (ICMR, 1972)
88 Harvard 50th percentile


84


80-- **


76-


72 -


68- *


64 -
(2)

60- *


56
3 7 11 15 19 23 27 31 35 39 43
Age (months) Rural Kerala









5




FACTORS DETERMINING
RATIONED RICE CONSUMPTION


In theory the subsidized food distribu-
tion system in Kerala is designed to supply
every individual with a fixed allowance of
approximately 1 kg. of rice a week at the
subsidized price.* In rural areas this rice
has to be purchased through a network of
fair-price shops-private or co-op retail
stores stocked periodically by the state
government. The amount a household
can purchase is fixed by the number of
unit members specified on its ration card.
Also based on the number of unit mem-
bers, households can likewise purchase
other subsidized items such as wheat or
wheat flour, sugar, and kerosene oil from
these shops.
In practice, there are many factors that
can affect consumption of the subsidized
food including, a household's knowledge
of and accessibility to fair-pricing shops,
supply conditions to and from the shops,
and the effective demand for the subsi-
dized foods. The latter is expected to be a
function of household incomes, the price
differential between the subsidized food
and its equivalent on the open-market,
and taste preference.
Consumption of rationed rice varied
widely among the three subregions in
the present study. The survey revealed a


pattern both in the proportion of partici-
pating households and in the level of sub-
sidized food consumption in the sub-
regions (Table 8). In subregion 1 all house-
holds participated, and their consumption
of rationed rice was the highest, with sub-
region 3 having the lowest figure on both
counts.
Itshould be noted that moving from sub-
region 1 to 3 means moving progressively
inland, with rapidly declining area under
rice cultivation, fewer employment op-
portunities and poorer all-around infra-
structure. The smaller supply of locally
grown rice and the predominance of lower
income households in the inland regions
would in fact suggest that demand for
ration rice on a per capital basis would be
highest and the price differential between
subsidized and open market rice would
be greater than near the coast even if
equal amounts of ration per capital were
allocated within the district. Instead,
practically all the households there de-
pend on the limited market for their rice
consumption. Unfortunately, information
on intradistrict distribution of rice to ra-
tion outlets is not available to see whether
or not that explains the lower consump-
tion of ration rice in some regions. Another


However, those classified as full producers of foodgrains are not eligible to purchase rationed rice. Only 3
percent of households with ration cards were listed as full producers in 1971-72 and hence not eligible to receive
rice on their ration cards. According to reported coverage figures of the retail outlets (which are likely to be somewhat
inflated) about 3.4 million households were eligible and received subsidized rice from the distribution system in
1971-72. The population reportedly reached was therefore about 80 percent (34, p. 43).









Table 8-Spatial differences in ration rice consumption

Subregions

1 2 3 Total
Sample


Number of households 24 12 7 43
Nonparticipating households
Number without ration rice 0 2 3 5
Percentage of households 0 17 43 12
Participating households
Ration rice calories (mean daily adult equivalent) 572 469 205 506
Standard deviation 245 103 115


possible explanation, also difficult to ver-
ify, is that a combination of factors en-
couraged leakages from the subsidized
food distribution system and that these
leakages were greater in some areas than
in others. The high price differential be-
tween subsidized and open market rice,
the reduced rice ration at the time of the
study, the smaller quantities of locally
produced rice available in the interior
areas provided better conditions for leak-
ages to exist here relative to the coastal
areas.
Most of the available evidence from
Kerala suggests that all the rice supplied
through fair-price shops is purchased by
consumers (8), suggesting the possibility
of a supply constraint.* At the time of this
study, the size of the rice ration was cut to
70 percent of the regular ration, so that
the supply constraint was probably even
more pronounced than usual. On the de-
mand side, purchasing power, knowledge
on use of the system, quality prefer-
ences,t and ability of higher income and


larger landowners to use ration rice as a
wage food are likely to be important. For
the present survey, it was possible only
to construct a simple model to test for the
statistical significance of some identifi-
able factors that determine rationed rice
consumption. The ordinary least squares
(OLS) equation used was:
RRE = a + BTY + yF
+ 81DR1 + 62DR2.
where RRE is monthly household expendi-
ture on rationed rice, TY is per capital total
household income per month, F is the
number of household members, and DR1
and DR2 are dummy variables to measure
intercept differences for subregions 1 and
2, respectively, from subregion 3. Results
of this regression equation are:
RRE = -27.4 + 0.03 TY + 6.3F
(4.8) (10.3)
+ 6.9 DR1 + 1.2 DR2
(1.8) (0.3)
t-ratios are in parentheses.
Family size shows up as the single most
important determinant of household pur-


* All of the rice is not necessarily purchased on ration cards alone.
t Most of the open-market rice sold may have a superior grain quality but is also polished like the rationed rice.








chases of rationed rice. This is to be ex-
pected since the number of family mem-
bers basically determines the issue of ra-
tion cards. But other factors are in prac-
tice also important.
Household income has a small, but sig-
nificant, positive effect on ration rice
expenditure. This positive correlation be-
tween ration rice consumption and income
is consistent with the low income levels
of the households and the large price dif-
ferential between the subsidized and
open market rice, which result in a high
income elasticity of demand at low in-
come levels. The association with per
capital incomes also implies that those
with somewhat higher incomes have
more favorable conditions for obtaining
their share or more than their share of
ration units, whereas, the poorest may
not have access to their full share. Some
evidence from Kerala does suggest that
the high demand for the ration may often
result in the poorest households mortgag-
ing their ration cards (9). This is probably
just one form of leakage, which could dur-
ing certain supply situations dilute the
distributive potential of such a subsidized
food system.
Coefficients of the subregional dummy
variables indicate differences between
subregion 1 and the other two subregions.
Some of the factors possibly responsible
for these differences have already been


discussed. At a macro level, access to sub-
sidized commodities often depends on
the level of organization, education, and
political pressures that a group can exert.
In the case of broad-based food subsidies
for consumers, such factors are also likely
to be important. But at the intraregional
level such influences would play a lesser
role, and access would be determined
more by supply and demand conditions
for the subsidized commodities. At the
household level, the regression equation
found incomes and family size significant
in determining ration access. Additional
information on education of either parent
failed to improve the explanatory power
of the equation and the education co-
efficients themselves were insignificant.
Furthermore, even though the pattern of
subsistence cultivation varies within the
district, it is probably safe to assume that
the taste preferences relevant in the selec-
tion of a food consumption bundle are
similar for the region. Therefore, the dif-
ferences observed are the result of de-
cisions made largelyon economic grounds.
The significance of differences in the
pattern of subsidized food consumption
- from the standpoint of analyzing both
food distribution policies and the em-
phasis needed in further research, will
become more evident after studying the
subsidy's impact on nutrition.










6



RATIONED RICE CONSUMPTION
AND NUTRITION



The results presented in this section
analyze the possible impact of rationed
rice consumption by households on:
1) levels of household nutritional in-
takes, and
2) measures of weaning-age child nu-
tritional status.
Dietary measures used at the household
level are caloric and protein levels and an
index constructed to reflect overall die-
tary diversity and quality. For the average
sample household, ration rice contributed
20 percent of both calories and proteins
in the diet. However, for the lowest in-
come households (for whom tapioca pro-
vides the bulk of calories), the potential
protein contribution of cheap rice be-
comes even greater. Because of the high
tapioca content of diets in the study house-
holds, it is important to consider the pro-
tein consumption measures together with
calories in evaluating the impact of ra-
tioned rice consumption. As noted earlier,
both caloric and protein consumption
levels were marginal. In the absence of
ration rice in the diets, increasing tap-
ioca consumption could maintain exist-
ing caloric intake levels, but protein con-
sumption would clearly become inade-
quate. It is not entirely clear, however,
that households consider tapioca and rice
as substitutes, at least in the short run.
Results in the next section suggest that
an increase in ration rice would not af-
fect tapioca consumption, would decrease
open market rice consumption, and would


increase total rice consumption given
present price relationships.
An index of dietary quality was also
constructed to reflect the consumption of
foods (other than the major caloric sour-
ces) that would be expected to improve
both the biological value of proteins and
the supply of essential vitamins and
minerals. From the weekly records of food
expenditures and consumption, the index
was derived as follows:

Index of overall dietary quality
5
SFi(Ei)/7
5

where i = 1 .... 5, are food categories, Fi
is the frequency with which the i category
was consumed that week, and Ei the ex-
penditure for it. The categories of foods
were milk and products; fish, meat and
eggs; pulses and legumes; vegetables;
and fruits: The average index for the six
weekly measures (each representing one
month) was used.
Percent standard weight for age, de-
scribed earlier, served as the indicator of
children's nutritional status. The average
of two observations was used for one
weaning-age child in each household.
The impact of ration rice on measures
of household nutrition was analyzed with
a set of equations with the format
Ni= f(TY, RSB) (2)
where Ni are four measures of household








nutrition used in this analysis: calories
per adult equivalent per day, protein per
adult equivalent per day, index of die-
tary quality for the household, and the
measure of weaning-age children's nutri-
tion status. TY is per capital total income
and RSB is a per capital estimate of the in-
come transfer derived by rationed rice
consumption. Both the income measures
are on a per month basis, and for all vari-
ables the mean of the six monthly mea-
sures is used in the analysis.
TSB, the value of ration rice subsidy,
is a quantity variable to which an aver-
age price differential is ascribed to derive
an estimate of the income transferred with
this ration. In this analysis
RSB = RRQ (OMRp RRp)
where RRQ is the quantity of rationed rice,
OMRp is the price of open market rice
and RRp is the price of rationed rice.
Using this price differential may inflate
the actual income transfer in terms of real
consumer purchasing power since trans-
action costs would certainly exist if the
ration is traded. The opportunity cost to
the state of providing the ration would
clearly be a function of the amount of
subsidy, RSB, but it would be a complex
relationship and no attempt will be made
here to resolve it. It must be emphasized
that for this analysis, the value (or cost)
of the subsidy per unit of ration rice is
academic-as long as it is applied uni-
formly for all households receiving it. Also
since the measure is averaged over six
months, the cost per unit of the subsidy
becomes a constant in time as well, elimi-
nating price variations from consideration.
The cost of the subsidy, therefore, acts
simply as a weight to the quantity of the
ration consumed and does not alter the


interpretation of either the consumption
or nutritional impact of access to the
ration.
Results of the OLS estimations for the
four measures of household nutrition are
given in Table 9, which gives the results
of a linear version of equations 2 and 3.
In equation,
Ni= f(FY, WY, RSB, DR1, DR2) (3)
household income is disaggregated into
farm and wage income components and
dummy variables are added to test for
regional effects. Here FY represents in-
come from farm production and WY in-
come from wages and remittances of
household members. DR1 and DR2 are
the intraregional parameters which test
differences for sites 1 and 2 respectively
from site 3. Ni are the same four house-
hold nutrition variables. The units for the
income measures are also the same as
before.
For all the measures of household die-
tary quality and nutrition, the amount of
ration subsidy had a significantly positive
impact. Improving these measures of nu-
trition with incremental rupees of ration
rice subsidy* was roughly six to ten times
more effective than improving them with
increments to the household's total in-
come packet. The wages income com-
ponent-the major income source for
these households-has a smaller impact
on all nutrition measures than own-farm
income (17). It is interesting to note that
the ration subsidy effect exceeds the own-
farm income effect on dietary variables
but compares with it in its effect on child
nutritional status. The subregional dum-
mies, in addition to testing for intra-
regional differences, were a useful test of


One unit of ration rice subsidy equals one Rs. of RSB per capital per month. Since the average price differential
between ration and open market rice during the study period was Rs. 2.40 per kilogram, one unit of RSB translates to
the purchase of 1/2.40 = 0.4 kilograms of ration rice per capital per month.









4 Table 9-Nutritional impact of household incomes and ration subsidy


(Aggregates for Six Monthly Measures)


Measures of Nutrition Equation 2 Equation 3
(Dependent Variable)

R2 Constant TYb RSBc R2 F Constant WYd FY' RSB DR1' DR2'


1. TOTCA 0.23 1,650.0 2.8 19.4 0.37 4.4 1,729.0 3.1 27.6 68.5 -577.0 -504.0
(calories per adult equivalent per day) (0.7) (1.2) (0.7) (2.2) (3.9) (2.5) (2.1)
2. TOTPA 0.52 22.4 0.16 1.54 0.55 9.2 22.3 0.10 0.57 1.56 0.41 2.15
(protein grams per adult equivalent CU per day) (2.1) (5.1) (1.2) (2.2) (4.3) (.08) (0.4)
3. Index of dietary quality 0.43 0.33 0.03 0.18 0.46 6.2 0.13 .03 .06 0.19 .01 0.76
(2.5) (3.7) (1.9) (1.3) (3.1) (.01) (0.9)

4. Nutritional status (child weight for age) 0.14 67.1 0.01 0.68 0.23 66.5 -0.05 0.46 0.64 0.7 0.1
(0.2) (2.3) (0.6) (1.9) (1.9) (0.2) (.02)


'Values in parentheses are t-ratios.
SPer capital total income per month.
Per capital ration subsidy per month.
'Per capital wages income per month.
"Per capital own-farm income per month.
' Regional covariance terms.








consistency for the income effects esti-
mators.t Only estimators for the calorie-
related nutrition measure (TOTCA)
changes with the covariance analysis,


suggesting that specification of this vari-
able is inadequate without a region-
related factor that independently affects
caloric intakes.


t This was done by first estimating equation 2 with and without the subregional parameters, and then substituting
the total income with an expenditure measure in equation 2 and repeating the estimation with and without the
subregional parameters to check coefficients for RSB.










7



RATION RICE
CONSUMPTION AND
HOUSEHOLD CALORIC INTAKE


This section attempts to further clarify
the role of ration rice consumption in
household caloric intakes and to deter-
mine a range for its possible net impact.
In addition, to understand the sources of
the observed caloric response and deter-
mine aggregate household demand impli-
cations for other foods, effects of ration
rice on consumption of other major
caloric sources are also investigated. The
functions to be estimated are
Ci = f(TY, RSB) (4)
and
Ci = f(TY, RSB, DR1, DR2) (5)
where Ci are total calories, tapioca calo-
ries, open-market rice calories, and total
rice calories, each on a daily/adult-equiv-
alent basis. TY and RSB are total income
and ration subsidy measures as before.
The estimation is repeated for each calor-
ic source with the addition of covariance
analysis using a subregional intercept
term or dummy. The results are presented
in Table 10. Using the double log form
yielded the best fit for all the caloric


equations. Marginal propensities to con-
sume (MPCs) for the double-log forms
and elasticities for the linear functional
forms for each of the caloric measures
used in relation to income and ration
subsidy have been computed at the mean
for each variable. The mean values were


Total income
(Rs. per capital per month)
Ration subsidy
(Rs. per capital per month)


Sample Mean
33.50

8.90


Total purchased calories
(daily adult equivalent) 1,900.00
Tapioca calories
(Rs. per capital per month) 830.00
Open-market rice calories
(Rs. per capital per month) 280.00
Total rice calories
, (Rs. per capital per month) 690.00
Results of the caloric equation are shown
in Table 10.
Calculations based on marginal pro-
pensity to consume calories provided by
ration subsidy,* indicate that at constant


The calculations were made using results of the double-log functional forms since they had superior R2 and
F values.
(i) For the six month study period the average price differential between ration and open-market rice was Rs.
2.40/kilogram.
(ii) Therefore, one unit RSB per capital per month is equivalent to the purchase of 1/2.40 kilograms of ration
rice = 0.4 kilograms. This provides 1,400 calories.
(iii) Converting the ration subsidy into adult equivalent terms, using the aggregate numbers of 6.5 resident
members and 5.0 adult equivalents (CU) per household, one unit RSB/adult equivalent is equivalent to
1,870 calories.
(iv) Therefore, percent net increment of calories per CU per month by the addition of one unit of RSB/adult
equivalent per month = MPC x 30
1,870 x100.
1,870The figures for C in the estimates of net caloric gain are 10.8 for the whole region and 21.6 within subregions
The figures for MPC in the estimates of net caloric gain are 10.8 for the whole region and 21.6 within subregions.










Table 10-Income and ration subsidy effects on household caloric
intakes
(Aggregates for six monthly measures)

Propensity to Consume Propensity to Consume
Additional Income-TY Subsidy Income-RSB
S Best
Equation Functional
MPCb IC MPC v7 Form


Total calories
Without subregional
dummy variables




With subregional
dummy variables




Tapioca calories
Without subregional
dummy variables

With subregional
dummy variables


Total rice calories
Without subregional
dummy variables


With subregional
dummy variables


Open-market rice calories
Without subregional
dummy variables


With subregional
dummy variables


(4) 2.8
(0.7)
5.1



(5) 5.9
(1.5)
10.9




(4) -9.9



(5) -3.0


0.05 19.4
(1.2)
0.09 10.8
(1.3)

0.11 39.5
(2.2)
0.19 21.6


(2.6)



-0.4
(2.4)

-0.12
(0.7)


(4) 6.3 0.29
(3.4)


(5) 2.8 0.13
(1.8)


(4) 8.2


1.0
(2.0)


(5) 4.9 0.6
(1.3)


-12.1



1.9





20.3



13.0





-14.3



-21.7


0.09 Linear


0.05 Double
(1.4) Log

0.2 Linear


0.10 Double
(2.8) Log



-0.13 Double
(1.7) Log


.02 Double
(0.26) Log



0.25 Double
(6.3) Log


0.16 Double
(4.4) Log



-0.46 Double
(1.9) Log


-0.7 Double
(3.3) Log


a Values in parentheses are t-ratios.
b MPC = Marginal propensity to consume calories with respect to the independent variables. Estimated at the mean
for double-log functions P y .
x
'T= Coefficient of elasticity. Estimated at the means for linear functions y .
x
dIntraregional covariance terms.









total incomes
1) for the whole region 17.5 percent of
calories provided by a unit of ration
subsidy show up as a net increase in
caloric intake;
2) within subregions, the net increment
in caloric intake is 35 percent of the
calories provided by a unit of ration
subsidy.
Why the difference? In part, because
patterns of dietary composition, especially
from the main calorie sources, vary even
within this fairly small area. Within a sub-
region, rationed rice consumption ap-
pears to replace open-market rice con-
sumption while the tapioca calories seem
to be unaffected. Since, as shown earlier,
access to ration rice is highest in the sub-
region with the highest open-market rice
consumption and lowest tapioca con-
sumption, the cross-regional analysis over-
explains a reduction in tapioca consump-
tion with ration subsidy.


This is a tentative explanation of what
is obviously a complex relationship. It
does, however, highlight the need to limit
the future analysis within ecological or
consumption zones, or at least to recog-
nize their possible influence when an-
alyzing food subsidy effects in larger
samples. As far as the present analysis is
concerned, the subregional dummy basic-
ally appears to be a stabilizing proxy for
omitted variables.
Therefore the estimate of a 35 percent
net caloric increment from ration calories
would more closely approximate the net
caloric efficiency of ration access. Total
rice consumption increases about 20 per-
cent per unit of ration rice consumed by
these households. Since these estimates
represent the response of predominantly
lower income households they do not
necessarily reflect changes in aggregate
demand.










8



DISCUSSION


Access to Subsidized Rice
Consumption


The question of ration access and use
can be analyzed from many different
vantage points-the biological, eco-
nomic, sociopolitical. The biological argu-
ment would predict that, ceteris paribus,
the better nourished would either be more
proficient in seeking access to the ration
or have the necessary energy available to
expend in the process of obtaining the
ration. A much more thorough under-
standing of exactly how an individual or
household gains access to the ration in
different circumstances is necessary to
develop an argument on biological
grounds. The tedium of the process of
obtaining ration rice does not by itself
suggest any clear hypothesis without evi-
dence to suggest the presence of extreme
caloric deficits relative to energy ex-
penditure.
It is important to recognize that while
planning for food needs of a population
justifies setting recommended caloric
levels commensurate with those needs-
based on "moderate" activity levels of the
entire population-much more caution
needs to be exercised in applying these
norms to derive the prevalence of mal-
nutrition (33). Estimates, such as 56 per-
cent of the population in developing
countries suffers severe caloric deficits
of 250 calories or more, paint the grue-


some picture of vast numbers of people
simply wasting away. They might be, if
all had the physique and level of energy
expenditure assumed in the derivation of
the requirements. In fact, however,
adaptation may occur to reduce the actual
levels of energy expenditure, although
the potential for reaching the levels pre-
dicted by the norms remains. For the
households studied the food supply was
probably inadequate for a productive and
active life based on present estimated re-
quirements. Other determinations of qual-
ity of life, such as provision of health ser-
vices and education, are important, but
clearly should not divert attention from
the search for ways to meet the basic food
needs of individuals. For example, as a
household level variable, education of
parents did not help to explain access to,
or consumption of, ration rice among
these households.
The way households behaved in con-
suming subsidized foods can probably be
explained best on economic grounds,
with the demand for a subsidized com-
modity determined by the quality of the
product, the price differential with the
open market, the level and distribution of
incomes, and income elasticities of de-
mand for the product. Based on these
factors, the greatest impact on consump-
tion by the poor can be achieved with the
least cost if the commodity is an im-
portant part of their diet, has a high in-
come elasticity at low income levels, and









has low or negative elasticities among
middle and upper income groups (20).
This has been empirically demonstrated
in the case of subsidized atta (wheat flour,
100 percent extraction) distribution in
Pakistan (29). In the present analysis,
even though ration rice was of an inferior
quality, it had a positive income elasticity
among the income groups that the sample
represents. It follows that consumers in
the middle income range derive a higher
absolute benefit than those in the lowest
income group. Two possible ways of im-
proving the targeting-both consistent
with reducing the cost of the subsidy are:
1) limiting the size of the subsidy to a
level that, for example, the two to three
lowest income deciles can be expected to
purchase, and 2) limiting the price dif-
ferential but selecting qualities and where
possible commodities with an inferior
prestige or preference value that are nu-
tritionally superior. In the case of rice,
parboiling might be advantageous in that
it offers the potential of lower milling
losses, superior nutritional content, and
in most cases* is probably considered
qualitatively inferior in consumption ex-
penditure patterns at middle and upper
income levels. New high-yielding varieties
(HYVs) offer another case in point since
they often do not meet the locally pre-
ferred qualities of flavor and texture.
There has been little explicit study on how
this set of preferences might interact with
the acceptance of the HYV seed by mar-
ket and subsistence producers, the dif-
ferential price levels for the varieties in
the open market, and the impact of price


supports and procurement on the quali-
ties of grain available and their distribu-
tional potential.
At the macro level, whether or not a
food subsidy program will work depends
on dominant social and political forces.
The broadest support for consumer sub-
sidies for basic commodities is likely to
coexist with the genuine participation of
the majority in the political process.t
Since the latter is a desirable goal and the
former one of its probably inevitable costs,
a much more thorough understanding is
needed of alternative ways to provide
such subsidies in ways that are cost-ef-
fective for the health of both the individ-
ual and the economy.


Impact on Nutrition

Why should the nutritional impact-
both in terms of diet and child nutritional
status-of subsidized rice be greater
than the impact of a cash subsidy (a sub-
sidy equivalent to the amount saved by
purchasing ration rather than open-mar-
ket rice)?
One explanation is that for these house-
holds the effective reduction in the price
of rice increases the quantity demanded
more because of the price effect (price
elasticity of demand) than because of the
income effect (income elasticity of de-
mand) as a result of the subsidy. Another
explanation is that the income effect does
predominate, but that the marginal pro-
pensity to consume food out of food sub-
sidy income is higher than out of aggre-


Traditionally parboiled rice has been the preferred grain in most rural areas of Southeast Asia, the Sri Lankan
wet zone and in Kerala. There is, however, some evidence to suggest that even in these areas, urban influence
serves to erode the traditional high regard for parboiled rice (11). In most of South Asia, parboiled rice appears to be
the inferior grain in terms of total consumer expenditures.
t Such subsidies are likely to receive broad support because they not only provide income transfers, but also help
to relieve the wage goods constraint for employment growth.









gate household income. That the marginal
propensity to consume differs by income
source has theoretical economic justi-
fication. Empirically, this paper has
shown that increasing income from own-
farm food production could result in sig-
nificant improvements in child nutrition
when aggregate wages income could not.
Also, for this sample of low income
households, where women participated
in the labor force, increments in wage in-
come (of both parents, but particularly
of the mother) were associated with im-
proved child nutritional status (17). If nu-
trition is seen as the result not just of a
household's consumption decisions but
of investment (in human capital) de-
cisions as well, then there is even greater
justification for expecting differences
because of income source. Just as farm
income might represent more stable or
permanent income, the allocation of
household resources by women might
tend to favor investment in children's
food consumption and nutrition, in both
cases resulting in a higher marginal pro-
pensity for nutritional improvement. It
is worth pointing out (Table 9, Equation 3)
that in the present analysis the estimated
beta coefficients for farm income and
ration subsidy income on child nutrition
are similar in magnitude, while the ag-
gregate income and wages income coef-
ficients are insignificant. In all cases, con-
sumption of subsidized food had a greater


impact on nutritional intake at the house-
hold level than an increase in income
equal to the food subsidy. Similar obser-
vations on the higher marginal propensity
to increase food consumption expendi-
ture from food subsidy income have been
made, both in low income (7) and high
income countries (1, 6, 19, 24, 27, 36).
Although such responses to food subsidy
income would be reflected relatively
more in nutritional intake and health
parameters in poor countries with lower
nutrition levels, both food stamp and
food subsidy programs in the United States
have improved intake of some nutrients
more than would be expected from an
equal income increment (18, 19).
A third possible explanation is an omit-
ted cash income source highly correlated
with the amount of ration rice purchased,
which is derived by selling the sugar ra-
tion eligibility that accompanies a rice
ration.* Some evidence of the sale for
cash of this luxury commodity entitlement
by low income households is available
(9). In effect, this mechanism makes a
cash transfer to low income households
and to traders in sugar as well. Where
legal sanctions prohibit resale to the same
trader (in Kerala it may be sold legally to
another agent), which is usually the most
convenient way to sell for a consumer
the sugar ration, there is little motivation
for either party to report it.


For each unit of ration eligibility in the public distribution system in Kerala, the holder is entitled to fixed
quantities of rice, wheat, sugar, and kerosene. Wheat and sugar consumption from the public distribution system
were, for different reasons, extremely limited for this sample of households.










9


CONCLUSIONS AND
IMPLICATIONS FOR
RESEARCH


Summarizing, household level data on
incomes, food consumption including
ration rice available, and child nutrition
from lower income households in rural
Kerala for six months in 1974 were an-
alyzed to study the role of access to sub-
sidized rice on levels of food consump-
tion and nutritional intake and status.
The analysis found:
1) These lower income households on
average suffered a deficiency of
both calories and protein in terms
of norms established by F.A.O. Rice
from the ration system contributed
one-fifth of both calories and pro-
tein in the household diet. Even
though the cost of a calorie from ra-
tion rice was nearly equal to the cost
of one from tapioca, one gram of
protein from ration rice cost only
one-fifth as much as one from tap-
ioca. Without ration rice, a net de-
cline in caloric and protein supply
would occur for these households
as they used that portion of con-
sumption expenditure to purchase
some tapioca but mainly open-mar-
ket rice.
2) During this period of foodgrain sup-
ply limitations, ration cutbacks and
a relatively high price differential
between the ration and open-market
rice, middle income groups had
more ration rice available than the
lowest income groups.
3) A large impact on consumption and


hence demand resulting from ration
rice availability is reflected in the
higher marginal propensity to con-
sume additional foods from the sub-
sidy income than from other income
sources. For these households a net
caloric increment of between 17 and
35 percent of ration rice calories
consumed was found to occur.
4) Ration rice consumption was also
positively related to child nutritional
status. This effect may be accounted
for by improvements in the caloric
and protein levels as well as in the
protein-per-calorie ratios in house-
hold diets.
Implications for predicting the health
and food consumption impacts of subsi-
dized food distribution systems can be
drawn only narrowly from this research
because of its limited coverage of geo-
graphical area, food consumption pat-
terns, selection of subsidized commodity,
and alternative types of subsidized food
systems. The results do suggest, however,
that there may be substantial potential,
both in terms of nutrition and via the in-
creased demand for foods, for linkages
with economic growth. In fact, there ap-
pears to be sufficient potential to merit a
systematic investigation of different sub-
sidized food distribution mechanisms.
Furthermore, cost considerations must
form an integral part of any assessment
of policy implications.














REFERENCES


1. Benus, J.; Kmenta, J.; and Shapiro, H. "The Dynamics of Household and Budget Allocation
to Food Expenditures." Review of Economics and Statistics 58 (May 1976): 129-138.
2. B. M. "Return to Free Market in Foodgrains?" Economic and Political Weekly 10 September
1977, pp. 1597-8.
3. Narain, Dharm. "The Working of a Dual Market System. A Short Note on Some of the
Issues Involved." Presented at the Colloquium on Agricultural Price Policy, Cornell Univer-
sity, February 1976.
4. Dorozynski, A. "Cassava-Solving the Toxicity Puzzle." 1 DRC Reports 7(March 1978):6-7.
5. Food and Agriculture Organization of the United Nations. Energy and Protein Require-
ments. Rome: F.A.O., 1973.
6. Food and Nutrition Service; U.S. Department of Agriculture. Impact of the Food Stamp
Program Upon the Food Marketing System of Puerto Rico. Part Two. San Juan: U.S.D.A.-
FNS and The Commonwealth of Puerto Rico, Department of Social Services.
7. Cavan, J. D. and Chandrasekera, I. S. "The Impact of the Public Foodgrain Distribution
on Food Consumption and Welfare in Sri Lanka," IFPRI Working Paper. Washington, D.C.,
1978.
8. George, P. S. "Public Distribution of Foodgrains--Welfare Implications and Effectiveness."
IFPRI Working Paper. Washington, D.C., 1978.
9. Gulati, L. "Rationing in a Peri-Urban Community-Case Study of a Squatter Habitat."
Economic and Political Weekly 19 March 1977, pp. 501-506.
10. Harriss, B.. "Beseiging the Free Market: The Effects of the Paddy-Rice Levy." In Green
Revolution? Technology and Chance in Rice Growing Areas of Tamil Nadu and Sri
Lanka. Edited by B. H. Farmer. Boulder: Westview Press, 1977.
11. Harriss, B. "Paddy Milling Problems in Policy and Choice of Technology." In Green
Revolution? Technology and Change in Rice Growing Areas of Tamil Nadu and Sri
Lanka. Edited by B. H. Farmer. Boulder: Westview Press, 1977.
12. Indian Council of Medical Research. The Nutritive Value of Indian Foods and the
Planning of Satisfactory Diets. Special Report Series, No. 42. New Delhi: I.C.M.R., 1966.
13. Indian Council of Medical Research. Growth and Physical Development of Indian Infants
and Children. Technical Report Series, No. 18. New Delhi: I.C.M.R., 1972.
14. Jacobson, H. N. "Current Concepts in Nutrition." The New England Journal of Medicine
297 (1977): 1051-53.
15. Jelliffe, D. B. The Assessment of the Nutritional Status of the Community. Geneva:
World Health Organization, 1966.









16. Kerala State Bureau of Economics and Statistics. "The Third Decennial World Census of
Agriculture, 1970-71 Report for Kerala State." Vol. 1. General Report. Mimeographed,
1973.
17. Kumar, S. K. "Role of the Household Economy in Determining Child Nutrition at Low-
income Levels: A Case Study in Kerala." Occasional Paper No. 95. Ithaca: Cornell
University, November 1977.
18. Lane, S. "Food Distribution and Food Stamp Program Effects on Food Consumption
and Nutritional Achievement of Low Income Persons in Kern County, California." Amer-
ican Journal of Agricultural Economics 60 (1978): 108-116.
19. MacDonald, M. Food, Stamps, and Income Maintenance. New York: Academic Press,
1977.
20. McCarthy, D. "Consumption Planning in Pakistan: Preliminary Analysis of Some Options."
Mimeographed. International Nutrition Planning Program, M.I.T. December 1975.
21. Mellor, J. W. "The Functions of Agricultural Prices in Economic Development." Indian
Journal of Agricultural Economics 23 (1968).
22. Mellor, J. W. Agricultural Price Policy and Income Distribution in Low Income Nations.
World Bank Staff Working Paper No. 214. Washington, D.C., September 1975.
23. National Sample Survey
24. Nelson, P. E., Jr. and Perrin, J. Economic Effects of the U.S. Food Stamp Program.
Agricultural Economics Report No. 331. Washington, D.C.: U.S.D.A. Economic Research
Service, July 1976.
25. PAHO. "Progress Report of the Inter-American Investigation of Mortality in Childhood."
Report presented to the ninth meeting of the Advisory Committee on Medical Research,
June 15, 1970.
26. Rao, K. S.; Swaminathan, M.C.; Swarup, S.; and Putwardhan, V. N. "Protein Malnutrition
in South India." World Health Organization Bulletin 20 (1959): 603-639.
27. Rees, R. B., Feaster, J. G. and Perkins, G. B. Bonus Food Stamps and Cash Income Supple-
ments-Their Effectiveness in Expanding Demand for Food. Marketing Research Report
No. 1034. Washington, D.C.: USDA Economic Research Service, October 1974.
28. Reutlinger, S. and Selowsky, M. Malnutrition and Poverty: Magnitude and Policy Options.
Occasional Paper No. 23. Baltimore: Johns Hopkins University Press for the World
Bank, 1976.

29. Rogers, B. L. and Levinson, F. J. "Subsidized Food Consumption Systems in Low-Income
Countries: The Pakistan Experience." Discussion Paper No. 6. International Nutritional
Planning Program, M.I.T., April 1976.
30. Sewell, John W. The United States and World Development: Agenda 1977. New York,
London: Praeger Publishers for the ODC, 1977.
31. Subbarao, K. "Market Structure in Indian Agriculture: A Study of Economic Efficiency of
Paddy/Rice Marketing System in West Godavari District, Andhra Pradesh." Ph.D. dis-
sertation, University of Delhi.








32. Sukhatme, P. V. "Incidence of Protein Deficiency in Relation to Different Diets in India:"
British Journal of Nutrition 24 (1970): 477-487.
33. Sukhatme, P. V. "Malnutrition and Poverty." Ninth Lal Bahadur Shastri Memorial Lecture,
Indian Agricultural Research Institute, New Delhi, 1977.
34. United Nations. Poverty, Unemployment and Development Policy-A Case Study of
Selected Issues with Reference to Kerala. New York: U.N., 1975.
35. Wall, Joh. "lndia-Foodgrain Management: Pricing, Procurement, Distribution, Import
and Storage Policy." Mimeographed. World Bank, South Asia Programs, November 1977.
36. West, D. A. and Price, D. W. "The Effects of Income, Assets, Food Programs, and House-
hold Size on Food Consumption." American Journal of Agricultural Economics 58 (1976):
725-730.
37. Wyon, J. B. and Gordon, J. E. The Khauna Study: Population Problems in the Rural Punjab.
Cambridge: Harvard University Press, 1971.






























































Shubh Kumar is a Research Associate at IFPRI. She joined the Institute in February 1978
after receiving a Ph.D. in international nutrition from Cornell University. Previously, she was on
the teaching staff of Punjab Agricultural University in India.




46






NOTES






NOTES







Intern nal Food Policy
Research Institute




Board of .Trustees


Sir John Crawford
Chairman, Australia

Ralph Kirby Davidson
Vice Chairman, U.S.A.

Ojetunji Aboyade
Nigeria

Nicolas Ardito Barletta
Panama

David E. Bell
U.S.A.

Norman E. Borlaug
Mexico

Ivan L. Head
Canada

Mohamed EI-Khash
Syria

Nurul Islam
Bangladesh

LucioG. Reca
Argentina

Roger Savary
France

Snoh Unakul
Thailand

Andrew Shonfield
United Kingdom

V. S' Vyas
India

John W. Mellor, Director
Ex Officio, U.S.A.














IFPRI PUBUCATIONS


MEETING FOOD NEEDS IN THE DEVELOPING WORLD: LOCATION
AND MAGNITUDE OF THE TASK IN THE NEXT DECADE

COMMODITY TRADE ISSUES IN INTERNATIONAL NEGOTIATIONS,
by Barbara Huddleston

RECENT AND PROSPECTIVE DEVELOPMENTS IN FOOD CONSUMP-
TION: SOME POLICY ISSUES

POTENTIAL OF AGRICULTURAL EXPORTS TO FINANCE INCREASED
FOOD IMPORTS IN SELECTED DEVELOPING COUNTRIES, by Alberto
Valdes and Barbara Huddleston

FOOD NEEDS OF DEVELOPING COUNTRIES: PROJECTIONS OF
PRODUCTION AND CONSUMPTION TO 1990

FOOD SECURITY: AN INSURANCE APPROACH by Panos Konandreas,
Barbara Huddleston, and Virabongsa Ramangkura









Reprinted
with permission of the
International Food Policy
Research Insitute




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs