EXPENDITURE ELASTICITY AND DEMAND PROJECTIONS FOR MAJOR FOOD ITEMS IN INDIA - A PANEL REGRESSION APPROACH

Kadakia International Journal of Research in Multidiscipline

ISSN: 2349 4875

Volume 1, Issue 1, June 2014 www.kijrm.com Economics | 117

EXPENDITURE ELASTICITY AND DEMAND PROJECTIONS FOR

MAJOR FOOD ITEMS IN INDIA - A PANEL REGRESSION

APPROACH

Nilesh B. Patel1

Gaurang Rami2

1Ph.D. Scholar, Department of Economics, Email: [email protected]

2Professor, Email: [email protected], Department of Economics

Veer Narmad South Gujarat University, Surat-395007, Gujarat, India.

Acknowledgement

We acknowledged deep sense of gratitude to Prof. M.B. Dave, Retired Associate

Professor, Department of Economics, VNSGU, Surat, for his guidance in clarifying

doubts and giving very useful suggestions in the draft version of this research paper. Prof.

Dave was kind enough to help us in data analysis and interpretations. We remain

indebted for his help, guidance and support.

Paper is presented in the 50

th Annual Golden Jubilee Conference of the Indian Econometric

Society (TIES) organized by Indira Gandhi Institute of Development Research (IGIDR), Gen. A

K Vaidya Marg, Goregaon (East), Mumbai - 400 065, India during December 22-24, 2013.

mailto:[email protected]

mailto:[email protected]


ISSN: 2349 4875


ABSTRACT:

During planning period the structure of Indian economy has undergone substantial

changes. There was also changes took place in the consumption pattern due to rise in

income levels and changes in the composition of income distribution. The significant

changes in the all sectors of the economy have a direct impact on the welfare of the

people. Knowledge of demand structure and consumer behaviour is essential for a wide

range of development policy questions like improvement in nutritional status, food

subsidy, sectoral and macroeconomic policy analysis, etc. The major objectives of the

present study are (i) to estimate expenditure elasticity of major food items in India for

rural and urban areas (ii) to make the demand and supply projections for major food

items in India (iii) to examine the gap between projected demand and supply of major

food items in India. To fulfill these objectives we have taken data from various rounds

(50th

round, 55th

round, 61st round and 66

th round) of NSSO. Since the data are panel in

nature, we have adopted the panel regression approach to estimate the income elasticity

(expenditure elasticity) and by using these estimated income elasticities the demand

projection has been made.

To determine the appropriate effects as fixed effect, random effect and pooled OLS

regression; we have used Joint test, Breusch-Pagan test and Hausman test. The fixed

effect panel regression model is applied in the case of cereals, pulses, milk, food oil,

vegetables and total food for rural as well as urban areas of India whereas random effect

panel regression is found suitable for MFC (Meat, Fish and Chicken) and Sugar for rural

areas of India and pooled OLS regression model found appropriate for urban areas of

India. On the basis of estimated expenditure elasticities of various food items in rural and

urban areas it was found that rural people are more responsive to change in the total

expenditure than urban people. However, in the case of milk the opposite situation is

found. The urban people are more responsive to milk consumption when their total

budget is changed. The projected data of demand and supply of various food items

implies that there will be a huge gap arises for cereals and vegetables in the future. In the

case of other food items the gap will be arise but not in sizable manner. This situation


ISSN: 2349 4875


suggest to policy makers that the focused should be made on the increase in the

production of cereals and vegetables by various ways such as increase in productivity of

land, utilization of land and other resources at efficient manner, adoption of modern

technology, multiple cropping patter etc. The expected gap between demand and supply

of various food items also useful to policy makers for design the policy regarding import

of these food items in future.

JEL Classification: C33, C53, O13

Key Words: Expenditure elasticity, food items, panel regression, fixed/random effect,

India

1. INTRODUCTION

Economic development has historically associated with structural changes in the national

economies. The most common structural changes that have been observed historically

have followed a sequence of shift from primary sector (agriculture) to secondary sector

(industry) and then to tertiary sector (services). These structural changes do not only

characterize economic development, they are also necessary for sustaining economic

growth. The neoclassical view that sectoral composition is relatively unimportant by

product of growth has been convincingly questioned by structural economists like

Kuznets, who have empirically demonstrated that growth is brought about by changes in

sectoral composition of national income. This is so both for the reasons of demand and

supply. Classicals like Fisher and Clark, basing their arguments on Engel‟s Law1,

thought that shift from agriculture to industry takes place as a result of low income

1 Engel's law is an observation in economics stating that as income rises, the proportion of

income spent on food falls, even if actual expenditure on food rises. In other words, the income

elasticity of demand of food is between 0 and 1. The law was named after the statistician Ernst

Engel. Engel's law doesn't imply that food spending remains unchanged as income increases: It

suggests that consumers increase their expenditures for food products (in % terms) less than their

increases in income. (for more details visit http://en.wikipedia.org/wiki/Engel's_law)

http://en.wikipedia.org/wiki/Engel's_law


ISSN: 2349 4875


elasticity of demand for agricultural products and high income elasticity of demand for

manufactured goods.

In India, after the economic planning the structure of Indian economy has totally changed

and there was also change in the consumption pattern of the Indian economy due to rise

in income levels and change in income distribution. The significant changes in the all

sectors of the economy have a direct impact on the welfare of the people. Economic

development results in increased levels of income and consumption; equally important, it

brings about a change in the socio-economic attitudes and even in the general outlook of

the people. A notable change in the pattern of consumption expenditure is expected due

to change in per capita income. The pattern of expenditure is a good indicator of the

economic status and the standard of living of the consumers and also shows the relative

importance of individual items in the consumption basket. Knowledge of demand

structure and consumer behaviour is essential for a wide range of development policy

questions like improvement in nutritional status, food subsidy, sectoral and

macroeconomic policy analysis, etc. An analysis of food consumption patterns and how

they are likely to shift as a result of changes in income and relative price is required to

assess the food security-related policy issues in the agricultural sector. This analysis is

based on a matrix of price and income elasticities of demand for food groups. In the short

run, with relatively inflexible production, changes in the structure of demand are the main

determinants of observed changes in market prices for non-tradable goods and of imports

and exports of tradable goods. In medium and long runs, the structure of final demand is

an important element of more complete models that seek to explain the levels of

production and consumption, price formulation, trade flows, income levels and

government fiscal revenues. Debates on the issue of food security in terms of the

country‟s self-sufficiency in production, future demand for cereals and other food items

as well as the ability of households to meet their calorie requirements are of important

policy relevance.

There were several methods used for estimating the expenditure elasticity. In the present

study have adopted the panel regression approach to estimate the income elasticity


ISSN: 2349 4875


(Expenditure Elasticity) and by using the income elasticity the demand projection has

been made by using demand projection model.


ISSN: 2349 4875


2. REVIEW OF LITERATURE

A review of available literature as related to the subject is an important and integral part

of any research study. A critical survey of the literature on the subject will help in

framing the aims objectives and methodology.

Stone Richard (1954) had analysed the pattern of demand for consumer's goods relating

to United Kingdom over the years 1920-1938 on the basis of annual data. Investigation

on different groups of consumer‟s expenditure, quantities bought and prices paid were

conducted. To analyse demand the study has applied Linear Expenditure System, which

is compatible with three conditions imposed on demand systems. i.e. (i) additivity, (ii)

homogeneity and (iii) symmetry. The analysis of a system of size commodity group,

among which the total of consumer's expenditure per equivalent adult has been divided,

is provided.

Saha Somesh (1980) had estimates the Engel elasticities for 101 items of consumption

separately for rural and urban India using NSS budget data. Iyengar's (1960-64) method

of estimation based on the use of generalized concentration curves had been used along

with method of weighted least squares for finding Engel elasticity of items. The estimate

seems to vary, though slightly, from one method to the other. However the ordering of

commodities on the elasticity scale is found to be approximately the same by all methods.

An inter-temporal comparison of elasticities over three different NSS rounds found the

Engel elasticity to be more or less stable across NSS rounds.

Abdulai et.al, (1999) had used Linear Approximate Almost Ideal Demand System

(LA/AIDS) to estimate food demand for India. The study discussed the need to use

demographic variables, such as, region, household size, education level, religion, and

seasonality in estimating food consumption in India. Past studies have used aggregate

household consumption data to estimate food demand in India due to non-availability of

micro-level data. The objective of this study was to show that factors other than price and

expenditure might be used to yield substantially greater precision in the estimation of

demand parameters. The study estimated separate food demand for urban and rural


ISSN: 2349 4875


population using household consumption survey data. The commodity groups used in the

study were: milk and milk products, cereals and pulses, edible oils, meat, fish, and eggs,

fruits and vegetables, and other foods. Demographic effects are incorporated in the model

allowing the intercepts in the budget share equations to be a function of demographic

variables. The results showed that all goods were normal since all expenditure elasticities

were positive. The commodity groups that had expenditure elasticities less than one in

rural and urban areas include cereal and pulses, edible oils and vegetables, and other

foods. The expenditure elasticities for milk and milk products were found to be greater

than one in rural and urban areas. All estimated compensated own price elasticities were

negative and ranged between -0.43 and -0.74 for rural areas and -0.46 and -0.74 for urban

areas. Interestingly, compensated cross price elasticities between cereals and pulses and

milk and milk production are found to be negative in rural area suggesting a

complimentary relationship between the food groups. However, in both rural and urban

areas, the cross-price elasticities were positive, suggesting a substitution effect between

groups.

Rolando Sammy Renteria, B.S. (2003) had analyzed the future supply and demand

situation for major grains (wheat, rice, and coarse grains) in India after taking into

account physical land constraints, urbanization and feed-livestock linkages. The study

used household survey data to estimate price and income responses of food demand

separately for urban and rural areas. The survey was conducted by our Indian collaborator

at the National Institute of Extension Management, Hyderabad, India during the period

from August 2000 to August 2001. The price and expenditure elasticities are estimated

separately for urban and rural areas. Study concluded that as expected, expenditure

elasticities for milk, meat, fish, eggs and fruits and vegetables are found to be high both

in the rural and urban areas. However, expenditure elasticities for major grains are found

to be relatively inelastic and slightly higher for rural and urban areas. All Hicksian own

price elasticities both in rural and urban areas are found to be negative and inelastic. Most

cross-price elasticities are found to be positive, suggesting that the food groups are

substitutes. Finally, the model is simulated with a set of exogenous assumptions to

project ten year supply, demand and trade of wheat, rice and coarse grains. The results


ISSN: 2349 4875


indicate that strong income growth and urbanization are expected to significantly change

the composition of the food basket. On average per capita cereal consumption is

projected to rise by around 8 percent from 166kg in 2002/03 to 179 kg in 2012/13.

Surprisingly per capita urban consumption which is 40 kg less than rural is projected to

increase by more than 24kg, mostly in wheat, in the next ten years. On the other hand,

rural per capita cereal consumption during the same period increased by a modest amount

from 166kg in 2002/03 to 179kg in 2012/13. Total cereal consumption is projected to rise

by 48 MMT, a 25 percent increase from the current consumption level.

On the basis of above reviews we can say that in most of the study the expenditure

elasticity had been estimating through OLS method and AIDS model. The AIDS model is

more popular for deriving the different types of elasticity. Very few studies have used

panel regression model to estimate expenditure elasticities using fixed and random effect

techniques, hence fixed and random effect model of panel regression model is used for

estimating the expenditure elasticity.

3. OBJECTIVES OF THE STUDY

This study has following major objectives;

(1) To estimate expenditure elasticity of major food items in India for rural and urban

areas.

(2) To make the demand and supply projections for major food items in India.

(3) To examine the gap between projected demand and supply of major food items in

India.

4. METHODOLOGY

Methodology is an important component of any research. Present study is descriptive and

analytical in nature. In order to fulfill the above mentioned objectives, an appropriate

methodology is adopted. Following are the details of sources of data, analytical tools and

techniques employed in this study.


ISSN: 2349 4875



ISSN: 2349 4875


Data:

The study utilizes the secondary data for evaluating and analyzing the specific objectives

of the study. Data on Monthly Per Capita Expenditure (MPCE) has been collected from

various round of National Sample Survey Organization (NSSO) published by GOI. The

data on physical quantities on selected food items were available only from 50th

round

(1993-94) onwards. In the present study the data on Monthly Per Capita Consumption of

various food items in monetary term had collected from 55th

(1999-2000), 61st (2004-05)

and 66th

(2009-10) rounds.

Estimation Method of Income/Expenditure Elasticity – Penal Regression Model

Since we have used data on similar variables for different rounds of NSSO, we find panel

regression model is best suitable to estimate income/expenditure elasticities. Panel

regression analysis is deals with two-dimensional panel data. Present study is based on

the secondary data on monthly per capita consumption expenditure of various food items

which had been collected from official website of NSSO. This data are usually collected

over time for the same states. The regression is run over these two dimensions.

A common panel data regression model is as follows;

Where PMCEFxst = Monthly Per Capita Consumption Expenditure on food item x for

state s……n and for the year t……..n; = Monthly Per Capita Total

Consumption Expenditure for state s……n and for the year t……..n. and are the

parameters of model.

The error is very important in this analysis. Assumptions about the error term

determine whether one can use fixed effects or random effects. In a fixed effects model,


ISSN: 2349 4875


is assumed to vary non-stochastically over s or t making the fixed effects model

analogous to a dummy variable model in one dimension. In a random effects model,

is assumed to vary stochastically over s or t requiring special treatment of the error

variance matrix.

Panel data analysis has three more-or-less independent approaches:

1. Independently pooled panels;

2. Random effects models;

3. Fixed effects models or first differenced models.

Independent Pooled Panel Regression:

Independent Pooled Panel regression model is assumed that the co-efficient of the model

is same within each unit (in our study this unit is each state). Therefore we can say that

Pooled panel regression model is simply like OLS model.

Fixed Effect Model:

Fixed effect regression model explore the relationship between predicator and outcome

variables within an entity (in our study this entity is state). Each state has its own

individual characteristics that may or may not influence the predicator variable (for

example being the gender of the person could influence on the consumption of various

food items). When we used the Fixed Effect Model, it assumed that something within the

individual may impact or bias t he predicator or outcome variable and we need to control

for this. This model gives the predicators‟ net effect. The fixed effects model is as

follows;

β0 + β1 1, + ………. + βk k, + γ1E1 + ….. + γnEn + ust


ISSN: 2349 4875


Where,

= is the Monthly Per capita Total Consumption Expenditure on Food item of

x for s state and t time.

= Monthly Per Capita Total Consumption Expenditure for state s……n and

for the year t……..n

βk = is coefficient food items xi….n

ust = is the error term

E1 = is the state n. since they are dummy you have n-1 states included in the model.

γn = is the coefficient for the dummies repressors.

The important assumption of this model is time invariant characteristics are unique to the

individual and should not be correlated with other individual characteristics. Each state is

different therefore the states‟ error term and the constant should not be correlated with the

others. If the error terms are correlated then fixed effect model is not suitable since

inferences may not be corrected and that time random effects model has been used.

Random Effect Model:

The random effect model has been used in this study when the differences across

different states have some influence on the dependent variable (Consumption expenditure

on different food items). An advantage of random effects model is that it can include time

invariant variables (i.e. gender, occupation ect.) In the fixed effects model these variable

are absorbed by the intercept. The random effects model is;

β0 + β1 , + ust + εst


ISSN: 2349 4875


Where, ust is explain the between state error and εst is within state error.

Random effects model assumed that the states‟ error term is not correlated with the

predicators which allows for time-invariant variables to play a role as explanatory

variable.

From the above models, which model is more suitable for our data set is selected by

using different statistical tests like Joint test, Breusch-Pagan test and Hausman test. The

joint test is used for the selection of panel regression model from pooled OLS model and

fixed effect model. The Breusch-Pagan test is used for the selection of panel regression

model from pooled OLS model and random effects model and Hausman test used for the

panel regression model selection between the fixed effects model and random effects

model. The null hypotheses of mention above test are as follows;

Joint Test:

H0 = the pooled OLS model is adequate, in favor of the fixed effects alternative

Breusch-Pagan test:

H0 = the pooled OLS model is adequate, in favor of the random effects alternative

Hausman test:

H0 = the random effects model is consistent, in favor of the fixed effects model

To run the above mention test and various panel regression models, we have used Gretl

open source software.

5. RESULTS

5.1 Expenditure elasticities of various food items in rural area:

In the following table the results of different test which have used for the selection of

panel regression model were given.


ISSN: 2349 4875



ISSN: 2349 4875


Table 5.1(a): Selection of Panel Regression Model for calculation of expenditure

elasticity of different food items – Rural Area

Food Items Joint Test

(P value)

Breusch-Pagan

test (P value)

Hausman test

(P value)

Fixed Effect/Random

Effect/Pooled OLS

Cereals 0.000 0.000 0.004 Fixed Effect

Pulses 0.000 0.000 0.018 Fixed Effect

Milk 0.000 0.000 0.056 Fixed Effect

Food Oil 0.000 0.000 0.010 Fixed Effect

MFC 0.000 0.000 0.434 Random Effect

Veg. 0.000 0.000 0.000 Fixed Effect

Sugar 0.000 0.000 0.432 Random Effect

Total Food 0.000 0.000 0.000 Fixed Effect

MFC: Meat, Fish and Chicken, Veg.: Vegetables

On the basis of above table we can say that only in the case of MFC (Meat, Chicken &

Fish) and the sugar consumption the random effects are found because the null

hypotheses test by Hausman test is not rejected. Therefore the random effects model has

been used for derived the expenditure elasticity for this food items.

On the basis of table 5.1(b) we can say that the expenditure elasticities of food items likes

cereals, pulses, milk, food oil, meat, chicken & fish, vegetables and sugar are 0.49, 1.01,

0.74, 1.01, 1.25, 1.31 and 0.78 respectively in rural area. For the pulses, food oil, meat,

chicken & fish and vegetables the expenditure elasticities have been noted to the greater

than one. So we can say that with increased one percent in total expenditure of rural

people, their expenditure on pulses, food oil, meat, chicken and fish and vegetables have

increased more than one percent. The lowest expenditure elasticities have been found for

cereals and highest for vegetables. The value of R2 has ranging 0.88 to 0.91 in the case of

fixed effects model. This value implies how much change in dependent variable due to

independent variable.


ISSN: 2349 4875


Table 5.1 (b): Expenditure Elasticity of Different Food Items in India (Rural Areas)

Food Items Intercept Elasticity R2

Cereals 1.563

(8.710)

0.487

(17.76)

0.90

Pulses -3.641

(-14.49)

1.01

(26.32)

0.91

Milk -1.217

(-1.982)

0.742

(7.916)

0.90

Food Oil -3.385

(-11.75)

1.007

(22.90)

0.88

MFC -5.072

(-8.773)

1.25

(15.06)

----

Vegetables -4.796

(-23.88)

1.31

(42.74)

0.94

Sugar -2.56

(-8.039)

0.775

(16.45)

---

Total Food -0.437

(-4.544)

0.973

(66.27)

0.98

Note: Figures in bracket indicate the t value



ISSN: 2349 4875


5.2 Expenditure Elasticities of Different Food Items – Urban Area

Table 5.2(a): Selection of Panel Regression Model for calculation of expenditure

elasticity of different food items – Urban Area


(P value)

Breusch-Pagan

test (P value)

Hausman test

(P value)

Fixed Effect/Random

Effect/Pooled OLS



Milk 0.042 0.169 0.053 Fixed Effect


MFC 0.763 0.385 0.660 Pooled OLS

Veg. 0.015 0.118 0.005 Fixed Effect

Sugar 0.077 0.173 0.970 Pooled OLS



In the case urban area, the polled OLS model is used only for the meat, chicken & fish

and sugar and the fixed effects model is applied for rest of the food items. For meat,

chicken & fish the null hypothesis is not rejected in joint test, Breusch-pagan test and

Hausman test. Therefore the Pooled OLS model is selected on the basis of joint test. The

same result is observed for sugar.

The expenditure elasticities for different food items in urban area are given in below

tables. The food items like cereals, pulses, milk, food oil, meat, chicken & fish,

vegetables and sugar were 0.26, 0.56, 1.21, 0.53, 0.93, 0.6 and 0.71 respectively. Only for

the milk, the expenditure elasticity is noted to the greater than one. The lowest

expenditure elasticity is found for cereals and highest for milk. Expenditure elasticities of

all food items had observed to statistically significant.


ISSN: 2349 4875


Table 5.2 (b) Expenditure Elasticity of Different Food Items in India (Urban Areas)

Food Items Intercept Elasticity R2

Cereals 2.952

(7.263)

0.264390

(4.614)

0.10

Pulses -0.758

(0.144)

0.563

(7.711)

0.639

Milk -4.243

(-3.336)

1.206

(6.72)

0.36

Food Oil -0.218

(-0.461)

0.528

(7.929)

0.67

MFC -3.077

(-2.227)

0.928

(4.766)

0.10

Vegetables -0.420

(-0.58)

0.634

(6.215)

0.54

Sugar -2.295

(-4.667)

0.705

(10.17)

0.33

Total Food 1.234

(6.20)

0.708

(25.21)

0.92

Note: Figures in bracket indicate the t value,


The value of R square is ranging 0.10 to 0.67. For the cereals and meat, chicken & fish

the value of R2

is very low (0.10), which indicate that only 10% changes in the

consumption of these food items is due to change in total expenditure. The low R2

value

is also observed for milk (0.36) and for sugar (0.33).


ISSN: 2349 4875


5.3 Expenditure Elasticities of Different Food Items (All India)

Table 5.3 (a) Selection of Panel Regression Model for calculation of expenditure

elasticity of different food items – All India


(P value)

Breusch-Pagan

test (P value)

Hausman

test

(P value)

Fixed Effect/Random

Effect/Pooled OLS



Milk 0.000 0.000 0.821 Random Effect


MFC 0.000 0.000 0.981 Pooled OLS

Veg 0.000 0.000 0.000 Fixed Effect

Sugar 0.000 0.000 0.972 Pooled OLS



In the above table the p-values of joint test, Breusch-Pagan test and Hausman test are

given. On the basis of this p-value we have selected the fixed effect model for cereals,

pulses, food oil, and vegetable and for total food. In the case of milk consumption the

random factors are affected and therefore the random effect model has selected.

However, for meat, chicken & fish consumption and sugar consumption the Pooled OLS

model is selected because the Hausman test show that the random effects model is better

than fixed effects model but join test implies that Pooled OLS model is quite better than

Fixed effects model.


ISSN: 2349 4875


Table 5.3 (b) Expenditure Elasticity of Different Food Items in India (Rural +

Urban Areas)

Food Items Intercept Elasticity Urban

dummy

R2

Cereals 2.26200

(8.957)

0.365

(10.26)

0.084

(3.153)

0.62

Pulses -3.45598

(-13.18)

0.942

(25.47)

0.270

(9.856)

0.80

Milk -3.77

(-6.15)

1.135

(11.25)

0.0167

(0.258)

Food Oil -2.940

(-12.80)

0.912

(28.12)

0.203

(8.403)

0.84

MFC -2.355

(-3.327)

0.835

(8.47)

-0.051

(-0.489)

Veg -3.524

(-11.84

1.07

(25.57)

0.270

(8.609)

0.79

Sugar -3.155

(-9.082)

0.824

(16.93)

0.288

(7.84)

Total Food -0.0474

(-0.461)

0.889

(61.21)

0.155

(14.32)

0.96



On the basis of above table we can say that the expenditure elasticity for different food

items like cereals, pulses, milk, food oil, meat, chicken & fish, vegetable and sugar were

0.37, 0.94, 1.14, 0.91, 0.84, 1.07 and 0.82 respectively for all India. These elasticities

have been found to be statistically significant at 0.01 and 0.05 significance levels. For

milk and vegetables, the expenditure elasticity are greater than one, which implies that

the consumption expenditure of these food items are more oriented to change in total

income. The lowest expenditure elasticity is recorded for cereals and highest for milk.


ISSN: 2349 4875


The coefficient of urban dummy is found to be significant for cereals, pulses, food oil,

vegetables and sugar. These coefficients are positive, which implies that the expenditure

elasticities of these food items in rural area were higher than urban area. The value of R2

is ranging 0.61 to 0.84 in the case of fixed effects model.

Table 5.3 (c) Comparison of Expenditure Elasticities of Rural and Urban Areas

Food Items Expenditure Elasticities

Rural Area Urban Area

Cereals 0.487

(17.76)

0.264

(4.614)

Pulses 1.01

(26.32)

0.563

(7.711)

Milk 0.742

(7.916)

1.206

(6.72)

Food Oil 1.007

(22.90)

0.528

(7.929)

MFC 1.25

(15.06)

0.928

(4.766)

Veg 1.31

(42.74)

0.634

(6.215)

Sugar 0.775

(16.45)

0.705

(10.17)

Total Food 0.973

(66.27)

0.708

(25.21)



The expenditure elasticity of cereals, pulses, food oil, MFC, vegetables and sugar had

found to be more than expenditure elasticity of these items in urban area. So we can say

that rural people are more responsive to change in the total expenditure than urban


ISSN: 2349 4875


people. However in the case of milk the opposite situation is found. The urban people are

more responsive to milk consumption when their total budget is changed.

6. DEMAND PROJECTION OF MAJOR FOOD ITEMS IN INDIA

The estimation of probable future demand for food items is essential for planner. It is

required to design major economic policies like food security, agricultural schemes,

import and exports of agricultural output etc. In this unit we have tried to project the

probable demand for major food items on the basis of projected population, future per

capita income growth and expenditure elasticity of these food items. The projected

probable demand of major food items has been calculated by using the demand projection

model given by International Agricultural Commodities and Trade (IMPACT). This

demand projection model was also used by Surbhi Mittal (2008). The demand projection

model is as follows;

Dt = d0 * Nt (1+y * e)t

Where,

Dt = household demand of a commodity in year t;

d0 = per capita demand of the commodities in the base year;

y = growth in per capita income; e is the expenditure elasticity of demand for the

commodity;

Nt = the projected population in year t.

For the calculation of probable future demand for major food items, we have to require

the data on projected population and average per capita income growth for projection

years. The data on projected population has been taken from the publication entitled “The

Future Population of India - A Long-range Demographic View” published Population

Foundation of India in 2007. The projected population in India is given in the following

table;


ISSN: 2349 4875


Table 6.1 Projected Population in India (In millions)

Year

Total

Population

Rural

Population

Urban

Population

% of Urban Population in

Total

2011

1203.71

(1.45)

812.51

(0.87)

391.21

(1.82) 32.50

2021

1380.21

(1.28)

869.02

(0.65)

511.19

(2.35) 37.04

2031

1546.16

(1.07)

831.65

(-0.45)

714.51

(2.85) 46.21

2041

1695.05

(0.88)

788.40

(-0.55)

906.66

(2.12) 53.49

2051

1823.52

(0.70)

753.53

(-0.46)

1070.01

(1.53) 58.68

Note: The Projected Rural and Urban Population are calculated on the base of estimated

urban population share in total population given in 2011 census provisional.

The Population Foundation of India had projected to 1203.71 million in 2011, which

increased and will reach to 1823.53 million in 2051. The decadal growth of the

population is assumed to be 14.55 % during 2001-2011, which declined over period of

time and came down to 7.05 % during the decade of 2041 to 2051. So we can say that in

future the population growth will decreased. It can be also seen that the urban population

will increases at increasing rate up to 2041 then it will increase at decreasing rate. It is

due to the high rate of migration (Urbanization) from rural to urban. It is estimate that

over a period of time the urban population share in total population will increase and

reach to 58.68% in total population in 2051.


ISSN: 2349 4875


Table 6.2 Alternative Per Capita Income Assumptions for Demand Projections (%)

Year Low Actual High

2011 2.05 4.05 5.55

2021 2.22 4.22 5.72

2031 2.43 4.43 5.93

2041 2.62 4.62 6.12

2051 2.80 4.80 6.30

The growth rates in per capita income under alternative scenario are worked out by

subtracting the population growth from income growth and then used for projecting the

per capita consumption of different food items.


ISSN: 2349 4875


Table 6.3 Projected Demands for Major Food Items in India

(Assumed to Alternative Per Capita Income Growth)

Projected Demand for Food Items (in MMT) Annual Growth Rate

PCI

Growth 2011 2021 2031 2041 2051

2011-

2021

2021-

2031

2031-

2041

2041-

2051

Cereals

Actual 370.77 436.17 503.04 566.48 623.80 1.50 1.329 1.120 0.92

Low 261.48 310.85 362.65 412.57 458.23 1.59 1.43 1.21 1.00

High 452.74 530.16 608.33 681.91 747.98 1.46 1.28 1.08 0.88

Pulses

Actual 50.03 59.34 69.06 78.40 86.93 1.57 1.408 1.191 0.98

Low 30.44 36.88 43.90 50.81 57.24 1.75 1.60 1.36 1.12

High 64.73 76.19 87.94 99.10 109.19 1.51 1.34 1.13 0.92

Milk

Actual 384.20 456.29 531.76 604.39 670.75 1.580 1.419 1.202 0.99

Low 228.10 277.30 331.24 384.56 434.26 1.77 1.63 1.39 1.14

High 501.28 590.54 682.15 769.26 848.12 1.51 1.34 1.13 0.93

Sugar

Actual 47.70 56.53 65.72 74.53 82.57 1.561 1.398 1.183 0.97

Low 29.58 35.75 42.44 49.01 55.11 1.72 1.58 1.34 1.11

High 61.30 72.11 83.17 93.67 103.16 1.50 1.33 1.12 0.92

Food Oil

Actual 49.36 58.53 68.11 77.30 85.68 1.567 1.405 1.189 0.98

Low 30.17 36.53 43.45 50.27 56.61 1.74 1.59 1.36 1.12

High 63.75 75.04 86.59 97.57 107.49 1.50 1.33 1.12 0.92

MFC Actual 33.67 39.90 46.40 52.63 58.31 1.562 1.400 1.184 0.97


ISSN: 2349 4875


Low 20.81 25.16 29.88 34.52 38.83 1.73 1.58 1.34 1.11

High 43.31 50.96 58.79 66.22 72.93 1.50 1.33 1.12 0.92

Veg.

Actual 519.52 616.75 718.45 816.26 905.60 1.577 1.416 1.198 0.99

Low 310.86 377.50 450.43 522.44 589.51 1.77 1.62 1.38 1.14

High 676.00 796.18 919.45 1036.62 1142.67 1.51 1.34 1.13 0.93


ISSN: 2349 4875


In the above table, the projected demand for various food items under the alternative per

capita income growth assumption is given. There are three alternative per capita income

growth assumptions are done. In all assumption the demand for various food items will

increased in future at all India level. But the rate of increased in demand for these food

items is noted to be declined in future. If we assumed that the per capita income will

remain increased by the actual rate, cereals demand will increased by 1.50%, 1.33%,

1.12% and 0.92% per annum during the period of 2011 to 2021, 2021 to 2031, 2031 to

2041 and 2041 to 2051 respectively. However, if assumed that the per capita income will

increase by lower rate, the demand for cereals will be increased by 1.59, 1.43, 1.21 and

1.00 per annum respectively. And if we assumed that the per capita income will be

increased by higher rate, the demand for cereals will increased by 1.46, 1.28, 1.08 and

0.88 percent per annum respectively. So we can say that the demand for cereals will

increased in projected time period in physical term but it will increase with diminishing

rate. The similar pattern in growth of the projected demand for various food items have

been reported in this study. If we assumed that per capita income will increase at low

rate, the demand will increased faster than actual growth and high growth rate

assumptions. However, when we assumed to high growth rate, the growth rate of

projected demand is less than low growth assumption and well as actual growth

assumption. So increasing rate of demand for various food items will be higher if the

economy grows at lower rate only. The projected demand for pulses, milk and vegetables

will increased at higher rate compared to the other food items. It is due to high elasticities

of demand for these food items. The growth rate of demand for various food items has

been noted to declined over a period of time can explained by the decreased in the

population growth rate in future. But when we considered the total demand of various

food items in quantity, it will be increased in future due to the increase in total population

in numbers in future.

7. SUPPLY PROJECTION IN INDIA

In the previous point we concluded that the demand for the various food items would be

increased all most for all items. However in future what will be happen is also depend on


ISSN: 2349 4875


supply situation of these food items. It is essential to predicate the future supply of

different food items for making the various strategies relating to food security in country.

The supply projection has been made using a straightforward approach. Supply

projections have been calculated assuming the yield growths to be same as in the past

decade. It is also assumed that further area expansion will take place. Supply projections

have been computed for the years 2021, 2031, 2041 and 2051 using the yield growth for

the most recent period of 2004-05 to 2011-12 and taking 20011-12 as the base year for

area and production.

The following formula has been used for supply projection*;

Yt = Y0*(1+r)t

Where,

Yt = Year of Projection of harvest area or yield of food items

Y0 = Harvest area or yield of food items in base year

r = average annual growth of harvest area or yield of food items

t = numbers of years under projection

After the calculation of projected harvest for food items area and yield of food items,

both projected values has been multiplied for calculate the projected production of

specific food items.

Table 7.1 Average Growth of Area under cultivation, Production and Yield of

Major Food Items During the period of 2005-06 to 2011-12

Items Area Production Yield

Cereals 0.4 3.6 3.26

Pulses 0.8 3.4 2.62

Sugarcane 3.8 4.9 1.36

Oil seeds -0.7 1.7 2.67

Vegetables 4.5 7.1 2.67

Source: Calculate from the various tables of Agricultural Statistics at a glance, 2012,


ISSN: 2349 4875


Directorate of Economics and statistics, Department of Agriculture and Cooperation.

It is clear from the above table that in last seven years, the area under cultivation,

production and productivity of food items like cereals, pulses, sugar cane, oil seeds and

vegetables had been increased. In the case of oil seeds the area under cultivation was

decreased but due to good productivity it is possible high production. The area under

cultivation for vegetables and sugarcane had been increased faster than other items i.e.

the area under cultivation for vegetables had increased annually by 4.5% and for

sugarcane it had increased by 3.8%. The productivity of the cereals was found to be

higher for cereals followed by oil seeds. Due to high increased in the cultivated area

under vegetables and sugarcane production the future production of these items estimated

to increases faster than other food items.

Table 7.2 Assumption of Maximum Land covered under harvesting for different

food items in future

Food Items Average

Growth Rate

Assumption

(000’ Hectors)

Cereals 0.4 150000

Pulses 0.8 35000

Sugarcane 3.8 8000

Oil Seeds -0.7 24474

Vegetables 4.5 16500

Source: Calculate from the various tables of Agricultural Statistics at a glance, 2012,

Directorate of Economics and statistics, Department of Agriculture and Cooperation.

The land is fixed factor of production, so when we estimate the projected harvest area for

different food items it should be keep in mind that the land cannot increased over a period

of time. Therefore, we assumed that at certain point the land for cropping of different

food items will become a fixed. We have assumed this on the basis of total available land

for agriculture and pattern of this land under the cultivation of different food items.


ISSN: 2349 4875


Table 7.3 Projected Supply of Major Food Items in India (in MMT)

Items

Scenerio-1

(Area Harvesting Growth is 3.26%)

Scenerio-2


2021 2031 2041 2051 2021 2031 2041 2051

Cereals 350.23 398.7 418.26 438.78 254.12 266.58 279.66 293.37

Pulses 30.86 40.85 52.81 68.26 22.09 28.55 36.91 47.71

Sugar 61.45 77.23 88.75 101.98 43.27 49.72 57.14 65.66

Food Oil 12.27 16.01 20.9 27.28 13.81 18.03 23.53 30.72

Vegetables 283.16 484.98 633.03 826.29 191.29 249.69 325.92 425.42

The projected supply of the different food items were given in the above table. This

projected is made by two scenarios, first assumed that area harvesting growth is 3.26%,

however at certain level the harvesting area were become a constant. The second scenario

is based on the assumption that there is no change in harvesting area for different food

items.

According to scenario one the supply of the cereals estimated to 350.23 million metric

tons in 2021, which will increased and reach to 438.78 million metric tons in 2051. The

pulses, sugar, food oil and vegetables supply is estimated to about 30.86, 61.45, 12.27

and 283.86 million tons in 2001 respectively, which will increases and reach to 68.26,

101.98, 27.28 and 826.29 million tons in 2051 respectively.

On the basis of second scenario, the supply of the cereals, pulses, sugar, food oil and

vegetables were estimated to 254.12, 22.09, 49.72, 13.81 and 191.29 million tons in

2021, which will increase and reach to 293.37, 47.71, 65.66, 30.72 and 425.42million

tons in 2051 respectively.

When, we compared the projected demand for various food items and supply of these

food items. It is observed that there will be wide gap arise in future. The availability of


ISSN: 2349 4875


the supply will be a smaller than demand for various food items. The following table

shows the gap between demand and supply of the selected food items in 2021, 2031,

2041 and 2051.

Table 7.4 Demand and Supply Gap

Food

Items

Scenerio-1

(Area Harvesting Growth is given table

no. 7.3 as 3.26%)

Scenerio-2


2021 2031 2041 2051 2021 2031 2041 2051

Cereals -85.94 -104.34 -148.22 -185.02 -182.05 -236.46 -286.82 -330.43

Pulses -28.48 -28.21 -25.59 -18.67 -37.25 -40.501 -41.49 -39.22

Sugar 4.92 11.51 14.22 19.41 -13.26 -16.00 -17.39 -16.91

Food Oil -46.26 -52.1 -56.4 -58.4 -44.72 -50.08 -53.77 -54.96

Vegetables -333.59 -233.47 -183.23 -79.31 -425.46 -468.76 -490.34 -480.18

The projected data of demand and supply gap of various food items given in the above

table shows that if we considered the scenario-1, excepting the sugar there will be deficit

in the availability of food items like cereals, pulses and food oil. However according to

the second scenario the sugar supply also become a less than its demand therefore there

will be deficit in availability of sugar also. In the case other food items the deficit noted

to be very huge according to the scenario-2.

8. CONCLUSION

We have adopted the three types of panel regression approach on the basis of different

test to estimate income elasticity (expenditure elasticity). The fixed effect panel

regression model is applied in the case of cereals, pulses, milk, food oil, vegetables and

total food for rural as well as urban areas of India. Therefore we can say that the

consumption expenditure of these food items had been not significantly affected by time-

invariant factors in both rural and urban areas. In the case of MFC (Meat, Fish and


ISSN: 2349 4875


Chicken) and Sugar for rural areas of India random effect panel regression model found

suitable and for the similar items pooled OLS regression model found appropriate for

urban areas of India.

The estimated expenditure elasticity of food pulses, food oil, MFC and vegetables had

been found to be greater than one in rural areas. So we can say that rural people are more

responsive to consumption of these food items when their total budget is changed. On the

other hand in urban areas the expenditure elasticity of milk was noted to greater than one,

which implies that urban people are more aware about the consumption of milk in their

food basket than rural people. The higher expenditure elasticity of food items indicate

that when their income level is increases faster rate the demand of these food items will

also increase greater proportion.

The projected data of demand and supply of various food items implies that there will be

a huge gap arises for cereals and vegetables in future. In the case of other food items the

gap will be arise but not at serious manner. This situation suggest to policy makers that

the focused should be made on the increased in the production of cereals and vegetables

by various ways like increase in productivity of land, utilization of land and other

resources at efficient manner, adopt the modern technology, multiple cropping patter etc.

The probable gap between demand and supply of various food items also useful to policy

maker to design the policy regarding import of these food items in future.

9. REFERENCES

1. Abdulai, A. (2002), “Household demand for food in Switzerland. A quadratic

almost ideal demand system”, Swiss Journal of Economic Statistics, 138, 1-18

2. Abdulai, A. and Aubert, D. (2004), “A cross-section analysis of household

demand for food and nutrients in Tanzania”, Agricultural Economics, 3 (1), 67-79

3. Abdulai, A., D. Jain and A. Sharma, (1999), "Household Food Demand Analysis

in India" Journal of Agricultural Economics, Vol. 50, 316-327


ISSN: 2349 4875


4. Ballino, Carlo (1990) „A Generalized Version of the Almost Ideal and Translog

Demand Systems‟, Economics Letters, 34, 127-129

5. Blundell, R., P. Pashardes and G. Weber (1993) „What Do We Learn about

Consumer Demand Pattern from Micro Data?‟, American Economic Review, 83,

570-97

6. Blundell, R.W. and Robin, J.M., (1999), “Estimation in Large and Disaggregated

Demand Systems: An Estimator for Conditionally Linear Systems”, Journal of

Applied Econometrics, 14, 209-232

7. Christensen, Laurits, Dale Jorgenson and Lawrence Lau (1975) „Transcendental

Logarithmic Utility Functions‟, American Economic Review, 65, 367-83

8. Deaton, A.S. and J. Muellbauer (1980) „An Almost Ideal Demand System‟,

American Economic Review, 70, 359-68

9. Dey, Madan Mohan (2000) „Analysis of Demand for Fish in Bangladesh‟,

Aquacultures, Economics and Management, 4, 63-81

10. Fisher, D., Fleissig, A.R. and Serletis, A., (2001), “An empirical comparison of

flexible demand system functional forms”, Journal of Applied Econometrics,

16(1), 59-80

11. Gould, W. and Villarreal, H.J., (2006), “An Assessment of the Current Structure

of Food Demand in Urban China”, Agricultural Economics, 34, 1-16

12. Gupta Anil, (1986), "Consumption Behavior in India- A Study of All India –

Consumption Estimates", Anmol Publications, Delhi, 12

13. Gupta Anil, (1986), "Consumption Behavior in India- A Study of All India

Consumption Estimates", Anmol Publications, Delhi, 12

14. Han, T. and T. Wahl. (1998), "China's Rural Household Demand for Fruits and

Vegetables" Journal of Agricultural and Applied Economics, Vol. 30(1), 141-150


ISSN: 2349 4875


15. Huang, J. and C. David. (1993), "Demand for Cereal Grians in Asia: the Effects

of Urbanization" Agricultural Economics. Vol.8, 107-124

16. Kumar, Praduman (2004) „Fish Demand and Supply Projections in India‟,

ICARICLARM Project, Division of Agricultural Economics, New Delhi: Indian

Agricultural Research Institute

17. Kumar, Praduman, (1997), “Food Security: Supply and Demand Perspective”,

Indian Farming, 4-9

18. Kumar, Praduman, (1998), “Food Demand and Supply Projections for India”,

Agricultural Economics Policy Paper 98-01, New Delhi, India: Indian

Agricultural Research Institute

19. Majumdar Amita (1980), “Consumer Expenditure Pattern in India: A

Comparison of the Almost Ideal Demand System and the Linear Expenditure

System”, Indian Statistical Institution

20. Majumdar Amita (1980), “Consumer Expenditure Pattern in India: A Comparison

of the Almost Ideal Demand System and the Linear Expenditure System”, Indian

Statistical Institution

21. Meenakshi, J.V. and R. Ray (1999) „Regional Differences in India‟s Food

Expenditure Pattern: A Completed Demand Systems Approach‟, Journal of

International Development, 11, 47-74

22. Mittal Surabhi (2010), “Application of The QUAIDS Model to the Food Sector in

India”, Journal of Quantitative Economics, Vol. 8 No.1, 43 – 54

23. Mittal Surbhi (2008), “Demand and Supply Trends and Projections of Food in

India”, Working Paper No. 209, Indian Council for Research on International

Economic Relations (ICSSR)


ISSN: 2349 4875


24. Mittal, Surabhi (2010), “Application of the QUAIDS Model to the Food Sector in

India”, Journal of Quantitative Economics, 8 (1), January

25. Mittal, Surabhi, (2006), “Structural Shift in Demand for Food: projections for

2020”, ICRIER Working Paper No 184

26. Mittal, Surabhi, (2007), “What Affects Changes in Cereal Consumption?”,

Economic and Political Weekly, 42 (5), 444-447

27. Mittal, Surabhi, (2008), “Demand-Supply Trends and Projections of Food in

India”, ICRIER Working paper No. 209

28. Moro, D. and Sckokai, P., (2000), “Heterogenous Preferences in Household Food

Consumption in Italy”, European Review of Agricultural Economics, 27(3), 305-

323

29. Muellbauer, J. and P. Pashardes (1992) „Tests of Dynamic Specification and

Homogeneity in Demand Systems‟, in L. Philips and L. D. Taylor (eds)

Aggregation, Consumption and Trade: Essays in Honour of Hendrik Houthakker,

Advanced Studies in Theoretical and Applied Econometrics, Kluwer Academic

Publishers, 55-98

30. Paul Bikas Sourabh (2011), “Food Preference and Nutrition in India”, University

of British Columbia, Vancouver, BC, Canada

31. Praduman Kumar et.al. (2011), “Estimation of Demand Elasticity for Food

Commodities in India”, Agricultural Economics Research Review, Vol. 24

January-June 201, 1-14

32. Rolando Sammy Renteria, B.S. (2003), “An Econometric Analysis of the Future

of Indian Food Supply and Demand”, A Thesis in Agricultural and Applied

Economics Submitted to the Graduate Faculty of Texas Tech University in Partial

Fulfillment of the Requirements for the Degree of Master of Science


ISSN: 2349 4875


33. Sadoulet, Elisabeth and Alain de Janvry (1995) „Demand Analysis‟, in

Quantitative Development Policy Analysis. Baltimore and London: The Johns

Hopkins University Press

34. Saha Somesh, (1980), "Some Further Estimates of Engel Elasticities for Rural and

Urban India", Sankhya, D, The Indian Journal of Statistics 1980,Vo1.42, Series-

D, Pts. 1 and 2, 127-150

35. Stone Richard, (1954), "Linear Expenditure Systems and Demand Analysis, An

Application to the Pattern of British Demand", The Economic Journal, Vol.64,

September, 511-527

36. Stone, Richard, (1954), "Linear Expenditure Systems and Demand Analysis, An

Application to the Pattern of British Demand", The Economic Journal, 64, 511-

527

37. Vaidyanathan, A, (1974), "Some Aspects of Inequalities in Living Standards in

Rural India", Poverty and Income Distribution, Statistical publishing Society,

Calcutta, 21S -241

EXPENDITURE ELASTICITY AND DEMAND PROJECTIONS FOR MAJOR FOOD ITEMS IN INDIA - A PANEL REGRESSION APPROACH

Documents