20 05 - Sae621826 - F
20 05 - Sae621826 - F
20 05 - Sae621826 - F
net/publication/293193514
CITATIONS READS
27 3,852
3 authors:
Hazrat Yousaf
Pakistan Institute of Development Economics
2 PUBLICATIONS 53 CITATIONS
SEE PROFILE
All content following this page was uploaded by Naveed Hayat on 24 August 2019.
Abstract
SE
This study analyzes the household food demand followed by projecting the future
level of demand of selected food commodity groups in Pakistan. It uses Pakistan
U
Panel Household Survey (PPHS) for the year 2010, conducted by Pakistan Institute
L
of Development Economics (PIDE). The linear approximation/almost ideal demand
IA
system (LA/AIDS) model is used to estimate the demand elasticities, while a
C
simple growth model is used for food demand projections. The empirical results
ER
reveal that food grains, pulses, ghee, sugar and vegetables are necessities, while
milk and meat are luxuries. Pulses and vegetables, ghee and meat, milk and sugar
M
substitutive relationship between different food items, such as pulses, meat and
O
sion in household demand for milk and meat products, but the consumption of
R
JEL: Q18, O4
T
O
Keywords
N
Introduction
Food demand analysis always remained an important issue among the econo-
mists. It differs across individuals, households, income, preferences, cultural tra-
ditions and local prices (Andersen & Watson, 2011). In the analysis of food
1
Pakistan Institute of Development Economics (PIDE), Islamabad, Pakistan.
Corresponding author:
Naveed Hayat, Department of Economics, Pakistan Institute of Development Economics (PIDE), Quaid-
i-Azam University Campus, P.O. Box 1091, Islamabad 44000, Pakistan.
E-mail: naveedhayat21@gmail.com
Hayat et al. 95
SE
production in the country. As a result, in the 1960s, green revolution in Pakistan
U
helped to increase the productivity of the agriculture sector. Especially, the total
productivity of wheat, rice and maize rose from 3,854,000 tons, 1,272,000 tons
L
and 531,000 tons to 7,123,000 tons, 2,372,000 tons and 567,000 tons, respec-
IA
tively. Before the green revolution, Pakistan used to import wheat, but the green
C
revolution raised the wheat production by 60 per cent. This increase in productivity
ER
and productivity growth rates were lower than population growth rates. From
M
1980 to 2010, the growth rate of the agriculture sector was steady between the
O
range of 2 per cent and 5 per cent, while the population growth rate was 2.6 per cent
C
during this period. From 2000 to 2005, both the agriculture growth and population
R
growth slowed down to 2.10 per cent and 1.9 per cent, respectively. In 2000 and
FO
2002, the agriculture sector showed a negative growth of –2.2 per cent and –0.1
per cent, respectively, whereas the population grew at 2.1 per cent and 1.8 per
cent, respectively. From the year 2007–08 and 2009–10, the agriculture sector
T
grew at 1 per cent and 0.6 per cent, respectively, whereas in the same period, the
O
population grew at 1.8 per cent and 1.7 per cent, respectively. Because of such a
N
high population growth and lower agriculture growth, the country remains a net
importer of several essential food items. Fluctuation in the production of food
commodities over time has turned the nation as food deficient (Zaheer, 2013).
Political unrest, militancy, natural disasters and energy crises had reduced the
income and employment level of the masses, which had further reduced purchas-
ing power or increased the poverty level in Pakistan. The rise in poverty or the
decline in purchasing power resulted in an increase in the percentage spending on
food. This means that other essential spending such as those on health and educa-
tion has been reduced. The share of household expenditure on food in the country
has increased to 61.6 per cent from 55.6 per cent in the year 2005–06 in the poor-
est group. An increase in percentage spending on food is directly related to market
prices and income level. This increase in food spending shows an increase in
poverty and consequently high vulnerability to food insecurity. In terms of
expenditure share on food, 28 per cent of the population was very poor, while
96 South Asia Economic Journal 17(1)
22 per cent had poor access to food. This means that 50 per cent of the population
has inadequate access to food, because of which many more people have dropped
down to the poor group (Suleri, 2009).
Changes in income, population explosion and industrialization in Pakistan are
also responsible factors for changing the structure of food demand. This may
shake the pattern of food demand both in present and in future. In view of the
importance of the food demand analysis and its expected implication for Pakistan,
this study provides answers to the following questions: What are the influencing
factors of household food demand in Pakistan? What is the future level of food
demand in Pakistan? The main objectives of the current study are to analyze the
household food demand in Pakistan and to project the future level of demand of
selected food commodity groups, such as food grains, pulses, ghee, milk, sugar,
meat and vegetables. Findings of the study will help the policy makers in devising
appropriate food policy through prices and demographic factors. Also, the food
SE
demand projections will help in formulating food policy, which, in turn, will
U
ensure food security in the country.
The remainder of the article is structured as follows. The following section
L
presents the review of literature. The ‘Data and Methodology’ section deals
IA
with data and the methodology of the study followed by the ‘Results and
C
Discussions’ section. Finally, the last section presents the conclusions and the
ER
policy implications.
M
M
Literature Review
O
C
Numerous studies have been conducted in relevance to the issue under considera-
R
tion, but we summarize some important ones here. Malik, Mushtaq and Ghani
FO
temporal differences in the expenditure pattern within the rural and urban sectors
N
of Pakistan. Burki (1997) used annual data of HISE from 1972 to 1992 to identify
changes in consumer preferences for eight food items. He used the generalised
axiom revealed preference model (GARP) and the linear approximation/almost
ideal demand system (LA/AIDS) model. The result of the study did not support
the GARP test, while it supported the LA/AIDS model for change in consumer
preferences. Farooq and Muhammad (1999) empirically investigated farm house-
hold consumption patterns. They used the concept of the almost ideal demand
system (AIDS) and found significant quantitative dietary impacts associated
with change in the age composition of farm households. Etzaz and Arshad (2007)
analyzed the household budget for Pakistan. They applied Quadratic Spline
Engle Equation on HIES 2000–01 data and found positive expenditure elastici-
ties for all the 22 commodity groups for both urban and rural households. Yousaf
and Khalil (2011) analyzed the households’ milk demand of Karachi. They used
the AIDS model and the HIES data for the year 2005–06. The results showed that
the household expenditure and demographic compositions by age were the main
Hayat et al. 97
determinants of household milk consumption. Zahoor et al. (2011) used the flexible
LA/AIDS model for the rural and urban households in order to find food demand
patterns for Pakistani Punjab. The study used HIES data for the year 2004–05.
Results of the study showed that the demand for all eight food commodity groups
was price inelastic. Sher, Ahmad and Safdar (2012) evaluated the household food
demand patters for various income groups in Pakistan. The study used Pakistan
Social and Living Standard Measurement Survey (PSLM) data for the year 2007–08
and the linear Engel’s curve. The household size and income elasticities were esti-
mated to explain the food consumption trends in Pakistan. Yousaf and Khalil (2012)
analyzed the household consumption patterns of Balochistan. They used the HIES
data for the year 2005–06 and applied the log linear and LA/AIDS models for the
analysis. The results of the study showed that the household expenditure on food
items was increasing at a decreasing rate. Khalil et al. (2012) analyzed the consump-
tion and expenditure patterns of seven food items for Pakistan, using the data of
SE
HIES for the year 2007–08 and LA/AIDS model. The results showed that Marshallian
U
own cost-price elasticities were negative for all food items, except for Mutton and
Fish in rural areas. Ahmad, Cheema and Saleem (2012) analyzed the food consump-
L
tion patterns of Pakistan at different income levels of the household at national as
IA
well as provincial levels. They used HIES for the year 1998–99. The results of the
C
study indicated that households in the lower income group had spend a higher por-
ER
tion of their incomes on necessities, while households in the higher income group
had spend a larger portion of their incomes on luxuries.
M
Most of these studies used HIES data sets for various years and different
M
models to analyze the food demand in Pakistan. Although the results of these
O
studies analyzed food demand and carried important policy implications, none of
C
these studies had provided projections of food demand. Therefore, the current
R
study makes an attempt to estimate the future level of food demand for Pakistan.
FO
T
Methodology
The estimation of Linear Approximation/Almost Ideal Demand System (LA/
AIDS) is carried out using a system of equations comprising household budget
shares for seven commodity groups. Thus, the estimated system consisted of a set
of seven budget share equations, that is, one budget share equation for each item
or commodity group. For ith commodity, the budget share equation used for
empirical estimation is,
where,
wi = Budget share for ith commodity group.
X = Per capita expenditure (R) on all consumption items included in the
SE
model.
P = Stones’ index estimated as ln P = S wj ln Pj1
U
Zh = No. of household members of type h.
L
h = 1, children (aged <6 years). IA
= 2, adolescents (aged 6–15 years).
= 3, adults (aged over 15 years).
C
where i, j = 1,2,3,…..N
M
wi = piqi / S pi qi (2)
R
group, respectively. piqi is expenditure on ith commodity group and Spiqi is total
expenditures (income).
T
pi = piqi / qi (3)
X = S piqi / n (5)
where, n denots Household size. The AIDS model satisfies the adding-up, homo-
geneity, and symmetry restrictions automatically. The adding-up requires Sgij = 0,
S bi = 0, homogeneity implies S gij = 0 and symmetry implies gij = gji. Based on
equation (1), the following uncompensated own-price (eii), cross-price (eij),
expenditure (Єi) and household age composition elasticities (fi) are estimated
from the parameter estimates:2
The values of sij are one and zero in the case of own price and cross-price elas-
ticity respectively. The compensated own and cross-price elasticities can be com-
puted by using the Slutsky equation in elasticity form:
SE
the following equation:
U
Ωih = [qih – bi ln (N+1/N)] * 100 (10)
L
IA
The bi parameter of the AIDS model determines the effect of a change in real
C
expenditure on the budget share of good i and whether this good is a luxury, a
ER
(X). For a necessary good, bi < 0 and the expenditure elasticity lies between zero
M
and one (0 < Єi < 1), wi decreases with increasing X. For an inferior good bi < –1
O
and the expenditure elasticity is smaller than zero (Єi < 0). In addition, it is pos-
sible to examine all complementary and substitutive relations between pairs of
C
Dt = d0 × Nt (1 + y × e)t (11)
T
O
where,
N
These projections have been based on constant price, 2 per cent growth of popu-
lation and various per capita income growths. Since the projected population
data of Pakistan are not available at any reliable source, therefore we projected
the population of Pakistan from the year 2011 to 2030 with the help of simple
compounding formula, taking 2010 as base year and 2 per cent as an average
population growth rate.5
SE
pulses are a minor consumption item group. The coefficients of variations for
U
Table 1. Descriptive Statistics
L
IA
Standard Coefficient of
Commodity Group Mean Deviation Variation
C
ER
Budget shares
Food grains 25.83 18.44 71.38
M
prices of various commodity groups showed the largest variation for the aggre-
gated price of milk category, while meat prices exhibited the least variation.
Regarding age composition, on an average every household is composed of 1.038
children, 1.189 adolescents and 4.652 adults and the average size of a household
is 7.509 members.
The system of demand equations in LA/AIDS model uses Seemingly Unrelated
Regressions (SUR) with homogeneity and symmetry restrictions imposed. To
preserve the adding-up restriction, one equation (vegetables equation) is omitted
from the model. The coefficient of this equation is obtained by imposing the
adding-up restriction. The estimated parameters of the preferred LA/AIDS model
are presented in Table 2 and its corresponding Mean Square Error (MSE) of
regression and adjusted R2 in Table 3, respectively. Out of the 84 parameters of
seven equations, 72 parameters are highly significant, while 12 parameters are
insignificant.6 The higher MSE of regression is observed for food grains while the
SE
lower MSE is observed for sugar. However, we have observed low MSEs for all
U
the seven regressions, which indicates the accuracy of our estimation techniques.
The highest adjusted R2 is observed for food grains, while the lowest adjusted R2
L
is observed for pulses and vegetables, respectively. It is observed that except food
IA
grains, all other food commodity groups have uncommonly low adjusted R2,
C
which is because we have used cross-sectional data set with a large number of
ER
observations here.
The estimated income/expenditure, uncompensated/Marshallian own, cross-
M
price and household age composition elasticities are presented in Table 4, while
M
of food grain, pulses, ghee, sugar and vegetables are less than one (Єi < 1), imply-
FO
ing that these items are necessities, whereas estimated income elasticities of milk
and meat are greater than one (Єi > 1), implying that these items are luxuries.
Comparing to Farooq and Muhammad (1999), who reported that pulses are neces-
T
sities and meat and milk are luxuries, and Zahoor et al. (2012) reported that veg-
O
etables and cooking oil are necessities and meat and milk are luxury food items.
N
Milk and meat are more income elastic than other food items, which implies that
when income rises then their demands also rise and it could be justified that due
to low purchasing power (poverty) in the country, the people were responding
more towards the consumption of these items as their income changes (Mudassar,
Aziz & Anwar, 2012). Except of meat, other six food commodity groups have
inelastic own-price (both uncompensated and compensated) elasticities, which
implies that they are integral items of household diet. Farooq and Muhammad
(1999) and Zahoor et al. (2012) reported inelastic own-price elasticities for all
food commodity groups included in their analysis. We are estimating both the
uncompensated and compensated cross-price elasticities. However, the uncom-
pensated/Marshallian own cross-price elasticity estimates provide the most
accurate picture of cross-price complements and substitutes. Therefore, we have
discussed only the cross-price effects of uncompensated elasticity estimates.
The negative value of uncompensated cross-price elasticities among pulses and
Table 2. Parameter Estimates of the Preferred LA/AIDS Model for Various Food Commodity Groups
Table 3. Mean Square Errors and Adjusted R2 of Preferred LA/AIDS Model for Various
Food Commodity Groups
SE
vegetables and among sugar and milk implies that they are gross complements7 in
U
consumption. Regarding the consumption of protein goods, the own-price elastic-
ity of meat is much higher than ghee, implying that households are relatively more
L
IA
responsive to change in the price of meat as compared to ghee. Whereas, their
uncompensated cross-price elasticities imply that they are gross complements in
C
consumption. Zahoor et al. (2012) also reported similar findings. The uncompen-
ER
pulses, meat and vegetables, respectively. Most of the household age composition
elasticities have positive signs but for milk and meat, it has negative signs. Farooq
M
demand, particular attention is focused on the sign of the coefficient of the respec-
R
tive age group variable. If, for example, a child is added to the household, ceterus
FO
paribus (including income), the child will place a specific additional demand on
the household’s consumption of food items (a hungry mouths effect). Since the
T
household in a monetary sense is now worse off, the child will reduce the house-
O
hold’s demand for (normal) food products (a real income effect).8 What is meas-
N
ured here is the combined impact of these responses. Thus, it is found that the
demand for all commodities except milk and meat increases with the addition of
a household member in each category (the hungry mouths effect outweighs the
real income effect). The negative signs of milk and meat for household composi-
tion imply that the households reduce the expenditure on these items with the
addition of members of various age groups. Moreover, adding a child, ceteris
paribus, reduces expenditure on milk and meat, respectively, of household income
(that is, the real income effect outweighs the hungry mouths effect), while increas-
ing expenses on food grains, pulses, ghee, sugar and vegetables. In short, we can
say that, when the household size increases by any age category then expenditures
on necessary food items group increases, while on luxury food items groups, it
decreases. This indicates the existence of poverty because increase in the house-
hold size reduces the household purchasing power, resulting in an increase in the
percentage spending on necessary food items, while reducing expenses on luxury
food items. Farooq and Muhammad (1999) also reported similar findings.
N
O
T
Table 4. Matrix of the Estimated Income and Uncompensated Own and Cross Price Elasticities of Demand for Various Food Commodity Groups
FO
R
With Respect to the Price of Household Composition
Commodity C
Group Income Food Grains Pulses Ghee Milk Sugar Meat Vegetables Children Adolescents Adults
Food grains 0.4580 –0.8213 0.0639 0.0486
O
–0.0049 0.1413 0.0298 0.0219 0.1031 0.2017 0.4078
Pulses 0.5872 –0.0798 –0.8092 0.0236 –0.1075
M 0.0205 0.0344 –0.0026 0.0848 0.1242 0.3177
Ghee 0.7531 –0.1418 0.0074 –0.8691 –0.1287
M –0.0492 –0.063 0.0613 0.2734 0.0598 0.2437
Milk 1.6003 –0.2419 –0.0056 –0.1398 –0.8311 –0.0642 –0.072 –0.1491 –0.0584 –0.1311 –0.261
ER
Sugar 0.6177 –0.0309 0.0424 0.0520 –0.1223 –0.7938 C 0.0224 –0.0200 0.0817 0.1050 0.3003
Meat 1.7167 –0.2527 –0.0268 –0.1373 –0.2598 0.0032 –1.018 –0.2068 –0.0780 –0.1121 –0.255
IA
Vegetables 0.7238 –0.0612 –0.0014 –0.0238 –0.0534 –5.6780 –0.025
L –0.6987 0.0497 0.1071 0.1711
Source: Computed by authors based on PPHS data of Pakistan for the year 2010.
U
SE
106 South Asia Economic Journal 17(1)
SE
Source: Computed by authors based on PPHS data of Pakistan for the year 2010.
U
L
Table 6. Percent Change in Household Income Spent on Various Food Commodity
IA
Groups Due to Change in Family Composition
C
ER
HH Composition
Commodity Children Adolescent
M
Source: Computed by authors based on PPHS data of Pakistan for the year 2010.
N
SE
Note: *In food demand analysis, we assume 1kg = 1l (Yousaf and Khalil, 2012).
U
Table 8. Base Year and Projected Population for the Year 2010 to 2030 (Million)
Year Population
L
Year Population
IA
2010 173.6 2021 215.5
C
2020 211.2
O
N
Source: Computed by authors based on Asian Development Bank (ADB) data set for the year
2010.
2011 to 2030. The details of base year and projected population are given in
Table 8. Base year and projected growth rate of per capita income/GDP is given
in Table 9. We have used the expenditure elasticities of various commodity groups
as derived in this study and given in table 5.
The total and per capita household demand of various food commodities from
the year 2010 to 2030 at 2 per cent population growth rate is given in Tables 10
and 11, respectively.9 It is observed that the total household demand for various
food items increases with the passage of time. The demand for food grains
increases from 18.6 million metric tons in the year 2010 to 36.4 million metric
tons in the year 2030. Household demand for pulses and vegetables rises from
16.7, and 21.7 million metric tons from the year 2010 to 35.3, and 49.8 million
metric tons in the year 2030, respectively. When we divide total household demand
108 South Asia Economic Journal 17(1)
Year Per capita Income Growth Year Per capita Income Growth
2010 2.69 2021 2.82
2011 1.42 2022 2.87
2012 2.06 2023 2.92
2013 2.02 2024 2.95
2014 2.54 2025 2.98
2015 2.97 2026 3.00
2016 2.90 2027 3.02
2017 2.82 2028 3.04
2018 2.68 2029 3.02
SE
2019 2.60 2030 3.05
2020 2.52
U
Source: Economic Research Service (ERS) Macro Economic Data set (2005).
L
IA
Table 10: Total Demand for Various Food Commodity Groups from 2010 to 2030
C
ER
by population, we get per capita household demand. From Table 11, we observe
rising trends in per capita household food demand from the year 2010 to 2030. Per
capita household demand for food grains, pulses, ghee, milk, sugar, meat and
vegetables rises from 107, 96.11, 27.6, 55.49, 85.7, 22.2 and 125 kg/per year in
2010 to 141.2, 137.1, 43.4, 144, 124.5, 61.6 and 193.3 kg/year in 2030, respectively.
Hayat et al. 109
Table 11. Per Capita Demand of Various Food Commodity Groups 2010 to 2030
(Kg/Year)
Commodity Groups
Year Food Grains Pulses Ghee Milk Sugar Meat Vegetables
2010 107.0 96.1 27.6 55.5 85.7 22.2 125.0
2015 114.5 104.8 30.8 70.0 93.9 28.4 139.0
2020 120.0 111.3 33.3 82.4 100.1 33.9 149.7
2021 123.2 115.1 34.7 90.2 103.7 37.3 156.0
2022 125.1 117.5 35.6 95.1 105.9 39.5 159.9
2023 127.1 119.9 36.6 100.5 108.1 41.9 164.0
2024 129.0 122.2 37.5 105.9 110.4 44.3 168.0
SE
2025 131.1 124.7 38.5 111.6 112.7 46.9 172.0
2026 133.1 127.1 39.4 117.5 115.0 49.5 176.2
U
2027 135.1 129.6 40.4 123.8 117.4 52.4 180.4
2028 137.2 132.2 41.4 130.5
L 119.8 55.4 184.8
IA
2029 138.9 134.2 42.3 136.1 121.8 58.0 188.4
C
Source: Computed by authors based on PPHS data of Pakistan for the year 2010.
M
M
Therefore, the food demand that has emerged from this empirical analysis sug-
O
gests that keeping prices constant when the population grow by 2 per cent per
C
annum, then per capita and total household food demand increases for the next
two decades. It may be concluded that household food demand has been primarily
R
This study attempts to analyze and project food demand in Pakistan. On one
hand, the pattern of food demand that has emerged from this empirical analysis
suggests that changes in the prices of milk and meat will bring major changes in
the diet of the households. On the other hand, a significant increase in the demand
of milk and meats can be expected following an increase in the household income.
An increase in household size, ceteris paribus, reduces the consumption of milk
and meat, but increases the demand for other food items. A change in the house-
hold age composition brings significant changes in the quantities of various com-
modities consumed. While, the food demand projections shows that per capita
and total household food demands would be approximately double for the next
two decades.
The results derived from this research have much policy relevance. The
expenditure elasticities of food items suggest that the nutritional requirements
will improve with the easy availability of food items. It may be generated through
110 South Asia Economic Journal 17(1)
the food support program like World Food Program (WFP), Benazeer Income
Support Program (BISP), Zakat and usher, etc. There exists a direct relationship
between household size and food consumption items because household size is
one of the most important determinants that significantly affect the household
food demand according to our findings, and therefore, various population control
measures may improve the standard of living of Pakistani households. The gov-
ernment can also consider the results of cross-price elasticities of food items in its
key decision regarding households. For instance, our results show that if a tax is
imposed on meat, households will substitute for meat into something that they
consider a good substitute. The policy makers can get further assistance from the
study findings while targeting household’s decisions. The projections for impor-
tant food items on various assumptions give the evidence of great responsibility
placed on the agriculture sector to produce food for domestic consumption. It is
because of high total and per capita demand for foods like food grains, pulses,
SE
ghee, milk, sugar, meat and vegetables, that the projected future demand for these
U
food items is quite big. The production in agriculture and dairy farming must be
directed towards increasing the supply of these foods items. Our estimated results
L
about food demand projection would be helpful in devising an appropriate food
IA
policy for the country.
C
The results of the study are satisfactory but further analysis may be of merit.
ER
Our data set provides information about major food commodities only and lim-
ited information about socio-economic factors. Therefore, our analysis focuses
M
only on the demand of major commodity groups while we are including one
M
factors such as education, occupation, region, household size, etc. in the model.
R
foods and ignore the indirect (animal) demand, which occupies a major share of
O
the total food demand. There is a need for further research to examine the direc-
N
Appendix
Theoretical Background
Demand analysis is a science of consumer/household choice/preferences among various
goods and services. The analysis of consumer demand is basically the act of analysis of
consumer preferences such as, how consumers choose to distribute their income among
Hayat et al. 111
different goods. Economic theory uses the concept of utility to define the level of satisfac-
tion that comes from the specific distribution of income among various commodities. There
are two basic problems of demand analysis: first is how to maximize utility and second is
how to minimize expenditures. The utility maximization condition is given by:
where, u is a utility function, x is total income of consumer and p and q are the prices and
quantities of kth commodities, respectively. On solving the first order condition for utility
maximization, we get Marsallian or uncompensated demand function:
qi = gi (x, P) (2)
where, P is the vector of commodity price. For a logarithmic utility function both income
and price elasticities can be calculated by taking the derivative of the Lagrangean function
SE
we get,
U
dlogqi = hi dlogx + Smij dlogpj (3)
L
where, hi is the income elasticity and mij are the uncompensated price elasticities. The fol-
IA
lowing conditions on the elasticities must hold,
C
ER
where, wj is the budget share. These two conditions of equation (2.4) are known as Engel
and Cournot aggregation, respectively and are known as the adding-up restriction. The
M
Solving the first order condition for expenditure minimization we get Hicksian or compen-
FO
qi = hi (u, P) (6)
O
N
Price elasticities derived from the Hicksian demand function are called compensated or
Slutsky price elasticities or Cournot price elasticities and is given as,
where, eij is the Slutsky price elasticity. Moreover, Hicksian demand function is equal to
Marshallian demand function at optimal utility level such as:
On the basis of Hicksian demand function, Deaton and Muelbauer (1980) developed
the Almost Ideal Demand System (AIDS) model. They derived the AIDS demand func-
tions in budget share form with the help of utility maximization and price-independent
generalized logarithmic (PIGLOG) preferences. The AIDS demand functions in budget
share form as:
For the present study, LA/AIDS is preferred because of its theoretical superiority, being
flexible in allowing but not requiring the general restrictions of demand theory to hold.
In addition, it permits a full range of commodities (complementary and substitute goods,
normal and inferior goods) to be analyzed.
Acknowledgement
Authors are grateful to two anonymous referees of this journal for useful comments. Views
expressed by the authors are personal. Usual disclaimers apply.
Notes
1. Introduced by Richard Stone in 1953.
2. See details in Farooq and Muhammad (1999).
3. This formula has been used by various researchers including Goyal and Singh (2004),
Mittal (2008) and Kumar, Joshi and Birthal (2009).
SE
4. PopFuture = Poppresent * (1+i)n
where: PopFuture = Future population, Poppresent = Present population, i = Population
U
growth rate, n = Number of year
L
5. The average population growth rate of the past ten year (from 2000 to 2010) estimated
IA
as 2% is consider for population projection.
C
6. The insignificant parameters are a7, g25, g35, g45, g52, g53, g57, q31, q32, q42, q71, q73 respectively.
ER
7. Two goods are said to be gross compliments if (dXi/dPi) < 0 and vice versa.
8. See details in Farooq and Muhammad (1999).
9. We report the estimated data for the cross-section years only due to space limitation.
M
M
References
O
Ahmad, A.U.F. (2010). The state of food security in Pakistan: Future challenges and coping
C
strategies. Paper presented at the 26th AGM and Conference of PSDE in Economics.
R
Ahmad, N., Cheema, A.R., & Saleem, A. (2012). Food consumption analysis in pakistan:
expenditure elasticities approach using HIES data. Interdisciplinary Journal of
Contemporary Research in Business, 4(4), 466–475.
T
Andersen, P.P., & Watson D.D., II (2011). Food policy for developing countries: The role of
O
government in global, national and local food systems. Ithaca: Cornell University Press.
N
Burki, A.A. (1997). Estimating Consumer Preferences for Food, Using Time Series Data of
Pakistan. The Pakistan Development Review, 36(2), 131–153.
Abid Qaiyum Suleri, S.H., de Franchis, Laura, Herbinger, W., Ramay, S.A., Ahmed, A.,
Anwar, A., & Salman, A. (2009). Food Insecurity in Pakistan. Pakistan: SDC Swiss
Agency for Development and Cooperation, SDPI Sustainable Development Policy
Institute, WFP World Food Programme Pakistan.
Deaton, A., & Muellbauer, J. (1980). An almost ideal demand system. The American
Economic Review, 70(3), 312–326.
Farooq, U.T.Y., & Muhammad, I. (1999). An investigation into the farm households
consumption patterns in Punjab, Pakistan. The Pakistan Development Review, 38(3),
293–305.
Kumar, P., Joshi, P.K., & Birthal, P.S. (2009). Demand Projections for foodgrains in India.
Agricultural Economics Research Review, 22, 237–243.
Malik, S.J., Mushtaq, M., & Ghani, E. (1988). Consumption patterns of major food items
in Pakistan: Provincial, sectoral and intertemporal differences 1979 to 1984–85. The
Pakistan Development Review, 4(II), 751–761.
Hayat et al. 113
Mudassar, K., Aziz, B., & Anwar, A. (2012). Estimating onsumer demand of major food
items in Pakistan: A micro data analysis. Pakistan Journal of Life and Social Sciences,
10(1), 53–58.
Mittal, S. (2008). Demand-supply trends and projections of food in India (Working Paper).
New Delhi: Indian Council For Research On International Economic Relations.
Muellbauer, A.D.J. (1980). An almost ideal demand system. The American Economic
Review, 70(3), 312–326.
Sher, F., Ahmad, N., & Safdar, S. (2012). Income and economies of scale effect on
household food demand pattern in Pakistan using PSLM data. Academic Research
International, 3(1), 50–56.
Yousaf, S., & Khalil, A.H. (2012). Analysis of consumption and demand elasticities for
food products in Balochistan. International Journal of Asian Social Science, 2(7),
1103–1122.
Zahoor ul Haq, H.N., Meilke, M., Muhammad, I., Khattak, A., Hasmi, A.H., & ur Rahman,
F. (2011). Food demand patterns in pakistani punjab. Sarhad Journal of Agriculture,
SE
27(2), 305–311.
Zaheer, D.R. (2013). Analyzing the performance of agriculture sector in Pakistan.
U
International Journal of Humanities and Social Science Invention, 2(5), 1–10.
L
IA
C
ER
M
M
O
C
R
FO
T
O
N