Pak J Commer Soc Sci

Pakistan Journal of Commerce and Social Sciences

2015, Vol. 9 (1), 159-170

Rural Urban Food Consumption Analysis in

Pakistan: Expenditure Elasticities Approach
Nisar Ahmad
Department of Economics, University of Sargodha, Bhakkar, Pakistan
Email: [email protected]

Muhammad Ramzan Sheikh

Department of Economics, Bahauddin Zakariya University, Multan, Pakistan
Email: [email protected]

Kashif Saeed
Department of Commerce, The Islamia University of Bahawalpur, Pakistan
Email: [email protected]

Abstract
The rural urban food consumption patterns are estimated and compared in Pakistan in the
present study. The Household Integrated Economic Survey (HIEs), 1998-99 data
published by the Federal Bureau of Statistics, Islamabad is used. The rural urban food
consumption analysis at household level is carried out by dividing the households into
five income groups at national and provincial levels. The expenditure elasticities of
different food groups are estimated across the different income levels to know the
differences among the rural urban household food consumption. The results of the study
indicate that rural and urban households have different food consumption patterns.
However, households in the low level income groups spend a larger fraction of their
income on wheat, pulses and vegetable while high income groups on rice, meat and fish,
milk and milk products both in rural and urban areas.
Keywords: Household Integrated Economic Survey (HIEs), rural food consumption
pattern, urban food consumption pattern.
1. Introduction
The households living in rural and urban areas make expenditure on different
commodities to attain utility and satisfaction. The expenditure on food commodities and
items are most important in the household behavior as food is basic nutritional ingredient
for every human being. Therefore, in the analysis of consumer behavior, the food
consumption is said to be the expenditure made by the consumer on different food
commodities, which he purchases to meet the daily food needs. It is observe that people
incurred expenditure on food items like wheat, rice, vegetables, fruits, meat and such
other kitchen items. The consumption of these food items of any household is generally
the function of income of household, prices of commodities, taste of the consumer and
other factors. Consumption function establishes relationship between consumption and
real disposable income of the consumer treating other factors constant.
 Rural Urban Food Consumption Analysis

Consumption expenditure on food commodities are used to indicate the living standard of
the household in a developing country like Pakistan. The food expenditures are also used
to measure the poverty level in the country. Consumption expenditure also incurred upon
the non-food items including electricity, clothing, furniture, housing, transport and
education; etc. to facilitate the several aspects of human life. Therefore, the analysis of
consumption and consumption patterns provides an insight into status of human resource
of a country. Therefore, consumption analysis of food at rural urban basis explains the
differences among regions in the country and provides the rational for future investment
planning decisions in the country. Such studies have attained special focus in recent years
in wake of globalization and concerns about food security.
2. Objectives of Study
Keeping in view the importance of food consumption, the main objective of the study is
to analyze food consumption pattern of rural urban households in Pakistan at national as
well as on provincial basis with different income groups. The study will be helpful to
compare and contrast the food consumption pattern among different regions in the
Pakistan. Expenditure elasticities of selected food commodities are calculated to know
the differences in consumption pattern of households belonging to rural urban areas of
Pakistan.
3. Review of Literature
Aguiar and Hurst (2005) show a comparison between household food consumption
patterns, expenditures and their status of unemployment and retirement in the United
States. The analysis is estimated by using two data sets that are the National Human
Activity Pattern Survey (NHAPS) and the Continuing Survey of Food Intake of
Individuals (CSFII) for the year 1989-91 and 1994-96. Permanent Income Hypothesis
(PIH) is employed to estimate income expenditure elasticities. The relationship between
household food expenditures, time spent on food production, and actual food
consumption is also examined. The changes in an individuals food consumption after
retirement is judged by the nutritional composition of the individuals diet, individual
categories of food consumption, consumption goods with an observable quality
component and consumption index. The consumption and expenditure analysis is based
upon the criteria of food consumption at home and away from home.
Syrovtka (2007) investigates the household demand pattern for various meat and meat
products by using the Household Expenditure Survey (HES) data of the years 1995-2000
in Czech. The exponential properties of Engels curve and income elasticities are
estimated by employing the Dynamic Engel model with the help of Ordinary Least
Square (OLS) technique. The highest magnitude of income elasticity is resulted in the
first quarter data of 1995 while lowest magnitude of income elasticity is resulted in the
fourth quarter data of 2000. It is also evaluated that meat and meat products are
considered to be luxuries for Czech households during the whole study period. The
overall meat demand pattern is only dependent upon the real level of household income.
The meat and meat products demand rises due to increase in the income level of the
households during the study period.
Fabiosa and Soliman (2008) determine the household expenditure pattern for four food
and non-food consumption heads in the urban and rural areas of Egypt. The main focus of
the study is to address the problem of food crisis and its alleviation measures. Household

160
 Ahmad et al

Income, Expenditure and Consumption Survey (HIECS) data of the eras 1999-2000 and
2004-2005 is used to analyze income-expenditure shares and elasticities by employing
Working-leser model. The HIECS data of 2004-05 give higher food consumption
elasticities as compare to the HIECS 1999-2000. There is not so much difference between
the expenditure shares of two survey periods and all the estimates of expenditure
categories are significant at one percent level in all equations. Lower income households
show a higher demand pattern for food rather than non-food commodities. It is also
observed that the rural households are more responsive to change in income for food
categories as compare to urban households.
Tey et al. (2009) compare the rural and urban household vegetables consumption trend
and determine the impact of household characteristics upon demand patterns of vegetable
in Malaysia. Semi Logarithmic functional form is used to estimate expenditure elasticities
applying Weighted Least Square (WLS) technique. The data is obtained from the
Household Expenditure Survey (HES) of the year 2004-2005. The expenditure elasticities
are found less than unity. Vegetable expenditures are found negatively related with the
household size in rural urban Malaysia and these expenditures are positively related with
the age of household members. Urban households show a higher quality demand pattern
for bulb, stem and processed vegetables whereas rural households make a quality demand
for fruiting, flowering and podded vegetables. The demand for leafy and salad vegetables
in both areas is positively affected by the income of the households.
Obayelu et al. (2009) assess the comparison of food consumption pattern in the rural and
urban areas of North-Central Nigeria. Household food consumption survey data of the
phase 2006-07 is used to analyze the household consumption differential for 15 food
items and nine types of meat commodities. The impact of household demographic and
socioeconomic characteristics is also determined and the overall estimation is done by
employing Double-Hurdle model. It is observed that the consumption of rice, oil, bread,
soft drink, sugar and milk is most common among urban households while yam and
cassava flour are the major consumed items of rural households. In case of meat items
beef, goat meat and fish are majorly consumed in the urban areas while bush meat and
ponma in the rural areas. The consumption of eggs, beans and such type of protein are
very lower due to high cost of these products and less awareness of the households about
its importance.
Oguoma et al. (2010) scrutinize the aggregate household demand system for edible oils in
the Imo State of Nigeria. Primary data is used in this study which is conducted from the
stratified random sample of 92 consumers of the two products across the three
agricultural zones in the state. Cobb-Douglass model is used to estimate own price, cross
price and income elasticities whereas the consumer preferences are determined by the
logistic techniques. The own and cross price elasticity of palm kernel oil is higher as
compare to the groundnut oil and it is not a close substitute for it. The demand for palm
kernel oil is reduces while for groundnut oil increase with an increase in consumers
disposable income so the income elasticity of palm kernel oil is lower as compare to the
groundnut oil. It is evaluated that if there is the reduction in the palm kernel price and
increase in the groundnut price then it will lead a strong competition between them for
revenue generation in the state.
Parpiev and Yusupov (2011) examine the household consumption pattern for seven food
and non-food consumption heads in rural and urban areas of Uzbekistan. Parametric and
161
 Rural Urban Food Consumption Analysis

nonparametric forms of Engel curves are estimated by using the Uzbekistan Regional
Panel Survey (URPS) data of the interlude 2005. The main focus of the study is to deal
with the problem of reported zero expenditures and to investigate the impact of
household characteristics on their demand by employing the Tobit estimation technique.
It is examined that the existence of economies of scale is most common in the household
consumption of education, clothing, transportation and health as compare to the shelter
and food consumption. The influence of household size on food is highly negative for
richest households as compare to poor and it is statistically insignificant for almost 25%
of poor population. The expenditure allocation for shelter is negatively significant for all
income groups; for health is positively significant for middle income classes and
positively significant for all the other commodities in all income groups.
4. Theoretical Framework
The research study on a specific topic is always carried out by exploring the related
theoretical structure. Therefore, the theoretical framework of the research study is
developed by explaining the associations of the concerned variables. These variables are
identified through interviews, observations, and literature survey. The building and
construction of the model building is important for the evaluation of the specific research
problem. Modelling the household consumption is much complicated and complex task
in the applied economics.
Innovative work related to modelling of household behaviour is completed first time in
the 1950s by Richard Stone. Cramer (1957) advance a neoclassical model integrating the
demand for durable and non-durable goods with the life cycle theory of Ramsey (1928),
Fisher (1930), Tinter (1938) and Modigliani and Brumberg (1955) under the guidance of
Stone. The essence of the model lies in the assumptions that the budget constraint is
linear and known with confidence and in efficiency-corrected units; new and used
durables are perfect substitutes. Stone (1954) first applied the stock adjustment model to
the demand for durables.
In most empirical studies on the households consumption patterns, which are based on a
single year cross-section data, it is customary to assume that for every commodity all the
households face the same prices and each individual faces the same utility function.
However, these assumptions are hardly true. In practice, cross section data does not
satisfy the conditions of single utility function. The preference ordering may very well
change from household to household depending upon the composition, e.g. number of
adults, children and females. There are various methods to take into consideration the
household composition effects in the Engel curve estimation.
Houthakker (1960) suggested that just deflate both demand and total outlay by household
size so that consumption is the same function of the budget and prices for all households.
Or other way is to use household size as an additional variable in the formulation of
Engel curve. Houthakker (1961) pointed out that coefficient of household-size represents
a combination of two effects, viz., the specific effect and the income effect. The
specific effect refers to the effect resulting from an increase in the need for various
commodities when the household-size increases. The increase in need is, however,
usually less than proportional to the increase in the household-size because of economies
of scale in the large household. The income effect refers to the effect when an increase
in the household-size makes the family relatively poorer in per capita income terms. If

162
 Ahmad et al

the specific effect dominates the income effect, the coefficient of the household-size is
positive; and negative otherwise. Another approach is to divide all expenditure by a
corresponding equivalence scale that converts the data to a need corrected basis.
In the light of Engle law, with an increase in income the share of expenditure on food in
total household expenditure tends to decrease, that on clothing, fuel and lighting remains
constant, and that on luxury goods increases. It is in this context that we should insert
Engels work, in which he stated the famous Engels law, the poorer a family, the
higher is the proportion of its total expenditure on foodstuffs; moreover, the wealthier a
nation, the smaller the proportion of foodstuffs in total expenditure. After the empirical
generalization of the Engels law; Schwabe, the director of the central statistical office
in Berlin, analyzed the relation between wages and rent for a sample of public employees
and added in 1868 that is known as Schwabs Law. This law states that the poorer a
family, the higher is the proportion of the income spent on rent.
Engel (Expenditure) elasticities are powerful research tool in the analysis of consumer
behavior, because informative results are concluded for the evaluation of consumption
patterns. In economics the deviation from proportionality of one variable with respect to
another variable is measured by elasticity. Thus, if expenditure on a certain item is
proportional to income (or total expenditure) then the income elasticity of demand, better
known as Engel elasticity, is unity. On the other hand, if expenditure on the item rises
more than proportionately with income, Engel elasticity is greater than one and if
expenditure on the item rises less than proportionately with income, Engel elasticity will
be less than one.
This concept is helpful in categorizing the commodities into different groups. The
commodity is considered to be a necessity if having elasticity less than unity, a need if
having elasticity equal to unity and a luxury if having elasticity greater than unity. The
elasticities of different commodities with respect to total expenditure are calculated and
results are compared across the commodities groups and income groups. These results
provide the guideline for future policy implication in respect of the management of the
demand and consumption of different commodities in the country.
5. Data and Methodology
The data of rural urban households is taken from Federal Bureau of Statistics, Islamabad.
This type of data is known as secondary source of data. For this purpose cross sectional
data compiled in the Household Integrated Economic Survey (HIEs), 1998-1999 is used.
The households are classified into five income groups to analyze food consumption of
rural urban households at national and provincial basis in Pakistan. These five income
groups (IG) include two low-income groups, IG 1 = Rs.1001-1500 and IG 2 = Rs.1501-
2000, two high income groups, IG 4 = Rs. 6001 to 7000 and IG 5 = Rs.7001 to above and
mean of all income groups i.e. IG 3. The included food items in the analysis are:
(i) Wheat and Wheat Flour (ii) Rice and Rice Flour
(iii) Pulses, Split and Whole (iv) Milk and Milk Products
(v) Meat and Fish (vi) Poultry
(vii) Fruits and Dry Fruits (viii) Vegetables
(ix) Tea, Coffee, and Soft Drink (x) Food (Aggregate)

163
 Rural Urban Food Consumption Analysis

A single utility function for all the rural urban households is assumed to estimate the
expenditure elasticities within each income groups, and that preference ordering does not
change across families. It is also assumed that for every commodity all the households
face the same prices because household survey does not provide information on market
prices. In estimating Engels curve, the consumption of different food commodities is
usually taken in terms of expenditure rather than quantities, because of the problem of
aggregation of heterogeneous items, and because expenditure takes care of the changes in
both the quantity and the quality of the goods consumed. The last assumption in this
regard is that expenditures are used as proxy for income.
The choice of an appropriate functional form in estimating the Engels curve is a matter
of great interest. As such, various functional forms, e.g., linear, double logarithmic, semi-
logarithmic, etc. are frequently used by the researchers (Gujarati, 1998). For the present
analysis, few functional forms are experimented. However, the results of hyperbolic form
are provided here as these results are found more appropriate and consistent with the
theory. The items for which a saturation level is expected a hyperbolic or reciprocal form
is better suited. i.e. Yi = i + i (1/ X). The slope and expenditure elasticity from this form
are: Slope = dY/dX = -i (1/X) Elasticity = dY/dX*Y/X = -i (1/XY). A hyperbolic
form has the property that there is an initial income below that an item is not purchased.
Here X is the per capita expenditure on commodity i and Y is the per capita total
expenditure (income) of the household. For calculating expenditure elasticities, the values
of the parameters i and i are estimated using method of Ordinary Least Squares (OLS).
The parameter i is used for the calculation of the slopes and elasticities.
6. Results and Discussion
The expenditure elasticities of different food items consumed by rural urban households
across the different income groups at national and provincial levels are reported in Table
1 to Table 5.
In so far as the results of expenditure elasticities of various food groups of rural urban
households are concerned, they are ranged to lie between zero and one for necessities and
greater than unity for luxuries. It can be argued on the basis of these estimates, the rural
urban households in different income groups exhibit different consumption patterns and
in general, they alter their consumption bundle both quantitatively and qualitatively in
response to changes in income.

164
 Ahmad et al

Table 1: Expenditure Elasticities, Pakistan Rural Urban

Rural Income Groups Urban Income Groups
Low IG Mean High IG Low IG Mean High IG
Commodity
Groups IG 1 IG 2 IG 3 IG 4 IG 5 IG 1 IG 2 IG 3 IG 4 IG 5
Wheat &
WF 0.53 0.46 0.37 0.36 0.26 0.12 0.09 0.05 0.06 0.04
Rice & RF 0.66 0.59 0.52 0.50 0.33 1.99 1.62 0.59 0.84 0.37
Pulses 0.79 0.66 0.51 0.46 0.36 0.44 0.35 0.21 0.27 0.15
Milk & MP 1.35 1.23 0.82 0.85 0.47 1.68 1.03 0.52 0.77 0.32
Meat &
Fish 5.45 4.63 1.84 1.78 0.93 4.87 2.35 0.71 1.13 0.41
Poultry 4.28 3.52 1.77 1.52 0.86 7.08 2.87 0.90 1.68 0.47
Fruits & D 3.46 2.97 1.65 1.52 0.82 3.52 2.19 0.71 1.12 0.40
Vegetables 0.47 0.42 0.33 0.31 0.24 0.64 0.42 0.24 0.32 0.17
Tea & Soft
Drink 1.70 1.28 0.90 0.85 0.56 1.78 1.27 0.52 0.83 0.31
All Food 1.14 1.02 0.72 0.68 0.45 1.29 0.74 0.42 0.60 0.27
Source: Calculated from HIES, 1998-99; WF = Wheat Flour, RF = Rice Flour & MP = Milk Product

A perusal of results show that in majority cases, the expenditure elasticities for lower
income households are high as compared to those for the higher income and the mean
income households. For example, expenditure elasticities for all rural urban Pakistan,
1998-99 in Table 1 confirms this statement for every commodity. In case of meat & fish,
the expenditure elasticity in the rural areas for lower income group is 5.45 and 4.63 as
compared to 1.78 and 0.93 for higher income groups and 1.84 for the mean income group
respectively.
In urban areas, these expenditure elasticities for lower income group are estimated as
4.87 and 2.35 as compared to 1.13 and 0.41 for higher income groups and 0.71 for the
mean income group respectively. Estimated expenditure elasticities of meat & fish are
observed high in the rural areas as compared to the urban areas. It indicate that the rural
households have less access to meat & fish and the people living in these areas are
nutrients deficient due to low levels of income. This statistics about the meat & fish
shows that expenditure elasticity is declining as moving from lower income groups to the
higher income groups in both cases of rural and urban areas. The same is to be found for
other food commodities in rural and urban areas.
As expected, the elasticity estimates for individual food items (in disaggregate terms)
provide some useful insight into the consumption level of various income groups. The
elasticities for most of food items found to be less than unity. Wheat, rice, pulses and
vegetables having elasticity less than unity across different income groups i.e. lower,
mean and higher income groups. The elasticity of milk and tea increases than unity only
in the lower income groups but not in the mean income group and higher income groups.
165
 Rural Urban Food Consumption Analysis

For example in Table 2, the expenditure elasticity of vegetables for rural households in
Punjab is found to be 0.98 and 0.81 in lower income groups, 0.72 in the mean income
group and 0.50 and 0.51 in the higher income groups.
Table 2: Expenditure Elasticities, Punjab Rural Urban
Rural Income Groups Urban Income Groups
Low IG Mean High IG Low IG Mean High IG
Commodity
Groups IG 1 IG 2 IG 3 IG 4 IG 5 IG 1 IG 2 IG 3 IG 4 IG 5
Wheat &
WF 0.99 0.82 0.65 0.47 0.43 0.71 0.64 0.19 0.19 0.09

Rice & RF 1.38 0.96 0.97 0.56 0.53 1.36 1.10 0.44 0.62 0.25

Pulses 1.47 1.22 0.82 0.54 0.56 0.60 0.43 0.25 0.30 0.16
Milk &
MP 1.38 0.86 0.98 0.71 0.52 2.02 1.27 0.59 0.82 0.33
Meat &
Fish 3.96 1.22 1.29 0.94 0.61 0.77 0.59 0.45 0.40 0.19

Poultry 4.77 3.56 1.78 1.15 0.80 4.37 2.98 0.93 1.63 0.44

Fruits & D 2.36 1.40 1.36 0.95 0.65 3.88 2.18 0.75 1.04 0.38

Vegetables 0.98 0.81 0.72 0.50 0.51 0.55 0.37 0.21 0.26 0.14
Tea & Soft
Drink 2.48 1.43 1.63 1.13 0.97 1.48 1.03 0.45 0.76 0.25

All Food 1.77 1.30 0.91 0.64 0.54 1.28 0.78 0.43 0.59 0.25
Source: Calculated from HIES, 1998-99; WF = Wheat Flour, RF = Rice Flour & MP = Milk Product

For urban areas, these elasticities are observed as 0.55 and 0.37 in lower income groups,
0.21 in the mean income group and 0.26 and 0.14 in the higher income groups. Keeping
in view the above discussion, it is concluded that all the food items in the highest income
group are of low elasticity i.e. less than unity. In the mean income group, the majority of
food items are having elasticity less than one and a small number of items having greater
than unity. High expenditure elasticities of poultry and fruit are found in the low level
income groups of both rural and urban households in Punjab. It explains the deficiency of
these food items for low level income households not only in the rural areas but also in
the urban areas of the Punjab.
The consumption pattern may vary among provinces of Pakistan and also in between the
rural urban areas of the specific province. It can be observed with the help of expenditure
elasticities of the commodities consumed in that specific region and province of Pakistan.
For instances, a perusal of expenditure elasticities for rural Sindh given in Table 3
indicates that the elasticity for rice is 0.21 and 0.11 in the lower income groups, 0.07 for
mean income household, and 0.02 and 0.01 in the higher income groups, respectively.
For urban areas, the elasticity for rice is 0.67 and 0.56 in the lower income groups, 0.38
for mean income household, and 0.59 and 0.27 in the higher income groups, respectively.

166
 Ahmad et al

Table 3: Expenditure Elasticities, Sindh Rural Urban

Rural Income Groups Urban Income Groups
Low IG Mean High IG Low IG Mean High IG
Commodi
ty Groups IG 1 IG 2 IG 3 IG 4 IG 5 IG 1 IG 2 IG 3 IG 4 IG 5
Wheat &
WF 0.48 0.34 0.27 0.24 0.20 0.39 0.33 0.46 0.25 0.16

Rice & RF 0.21 0.11 0.07 0.02 0.01 0.67 0.56 0.38 0.59 0.27

Pulses 0.53 0.47 0.31 0.29 0.23 0.08 0.14 0.07 0.11 0.06
Milk &
MP 0.73 0.62 0.42 0.36 0.27 1.51 1.16 0.52 0.83 0.36
Meat &
Fish 1.54 1.12 0.74 0.69 0.46 4.82 2.12 0.75 1.24 0.52

Poultry 2.24 1.89 1.11 0.84 0.58 4.32 3.59 0.90 1.17 0.55

Fruits & D 9.45 6.36 1.49 1.29 0.81 1.90 1.82 0.74 1.34 0.48

Vegetables 0.07 0.06 0.05 0.05 0.04 0.38 0.35 0.23 0.31 0.18
Tea & Soft
Drink 2.42 2.08 0.96 0.84 0.62 0.81 0.76 0.46 0.79 0.31

All Food 0.81 0.75 0.45 0.70 0.32 1.14 0.76 0.45 0.70 0.32
Source: Calculated from HIES, 1998-99; WF = Wheat Flour, RF = Rice Flour & MP = Milk Product

It explains that urban households have greater elasticities as compared to the rural
households in case of Sindh. Therefore, the rural people are more inclined towards the
consumption of the rice. It may be due to rice is their staple food and they have develop
the taste for the consumption of the rice. This range is also to be found very much lower
as compared to other provinces of Pakistan.

167
 Rural Urban Food Consumption Analysis

Table 4: Expenditure Elasticities, NWFP Rural Urban

Rural Income Groups Urban Income Groups
Low IG Mean High IG Low IG Mean High IG
Commodity
Groups IG 1 IG 2 IG 3 IG 4 IG 5 IG 1 IG 2 IG 3 IG 4 IG 5
Wheat &
WF 0.36 0.25 0.41 0.19 0.11 0.45 0.20 0.19 0.25 0.14

Rice & RF 2.89 1.71 1.25 1.22 0.79 2.60 2.23 0.75 1.04 0.46

Pulses 0.22 0.15 0.15 0.14 0.13 0.49 0.26 0.26 0.29 0.19
Milk &
MP 0.93 0.82 0.70 0.65 0.47 1.38 0.65 0.57 0.93 0.34
Meat &
Fish 3.99 3.00 2.36 2.44 1.23 2.61 1.13 0.84 1.59 0.45

Poultry 2.74 2.14 1.57 1.54 0.88 4.36 2.54 1.12 2.79 0.54

Fruits & D 5.82 2.34 1.80 1.76 1.02 2.15 1.37 0.79 1.67 0.42

Vegetables 0.98 0.59 0.57 0.54 0.44 0.58 0.55 0.32 0.48 0.21
Tea & Soft
Drink 0.75 0.57 0.49 0.46 0.37 0.32 0.22 0.20 0.31 0.14

All Food 0.98 0.72 0.63 0.61 0.44 1.03 0.87 0.46 0.72 0.29
Source: Calculated from HIES, 1998-99; WF = Wheat Flour, RF = Rice Flour & MP = Milk Product

For the analysis of wheat consumption, it is observed that expenditure elasticity is highest
in Balochistan and is lowest in the NWFP for both rural and urban households. The
results about these elasticities are reported in Tables 5 and 4 respectively. In Table 5, the
estimates of wheat elasticity for rural households of Balochistan are 1.79 and 1.20 for
lower income group, 0.79 for mean income group; and 0.88 and 0.64 for higher income
groups. For the urban households, the estimates of the elasticities are 0.53 and 0.36 for
lower income group, 0.29 for mean income group; and 0.24 and 0.22 for higher income
groups.

168
 Ahmad et al

Table 5: Expenditure Elasticities, Balochistan Rural Urban

Rural Income Groups Urban Income Groups
Low IG Mean High IG Low IG Mean High IG
Commodity
Groups IG 1 IG 2 IG 3 IG 4 IG 5 IG 1 IG 2 IG 3 IG 4 IG 5
Wheat &
WF 1.79 1.20 0.79 0.88 0.64 0.53 0.36 0.29 0.24 0.22

Rice & RF 2.88 2.72 2.33 1.93 1.05 1.26 0.94 0.97 0.83 0.76

Pulses 0.76 0.64 0.51 0.53 0.46 1.48 0.91 0.82 0.87 0.83
Milk &
MP 0.25 0.19 0.17 0.17 0.13 0.89 0.27 0.51 0.82 0.62
Meat &
Fish 2.19 1.50 0.97 1.04 0.75 4.56 2.24 1.17 1.24 0.91

Poultry 3.51 2.82 1.30 1.27 1.02 3.26 2.29 1.26 1.46 0.92

Fruits & D 3.54 2.81 1.33 1.44 0.95 7.27 3.83 2.56 2.57 1.94

Vegetables 0.08 0.06 0.05 0.06 0.04 2.57 2.04 1.36 1.34 1.13
Tea & Soft
Drink 1.26 0.98 0.68 0.76 0.58 0.95 1.76 0.83 0.90 0.76

All Food 0.87 0.84 0.59 0.65 0.48 1.96 1.67 1.14 1.23 0.98
Source: Calculated from HIES, 1998-99; WF = Wheat Flour, RF = Rice Flour & MP = Milk Product

Whereas in Table 4 for comparison with NWFP, wheat elasticities for rural households in
NWFP are 0.36 and 0.25 for lower income group, 0.41 for mean income group and 0.19
and 0.11 for higher income groups. For the urban households, these elasticities are 0.45
and 0.20 for lower income group, 0.19 for mean income group and 0.25 and 0.14 for
higher income groups.
For rural urban food consumption comparison, it is found that rural NWFP households
have highest elasticities for meat & fish in the lower income groups. These elasticities are
3.99 and 3.00 for rural NWFP households in the lower income groups as compared to all
other provinces. In Sindh for the lower income groups, the elasticity for meat & fish 1.54
and 1.12 respectively; in Punjab are 3.96 and 1.22; in Balochistan are 2.19 and 1.50
showing that people of low income groups in rural areas of NWFP have less consumption
of meat & fish as compared to other provinces. For lower income groups in urban areas,
the elasticity for meat & fish is found highest in Sindh. These elasticities are 4.82 and
2.12 for urban Sindh households in the lower income groups as compared to all other
provinces. In Punjab for the lower income groups, the elasticity for meat & fish are 0.77
and 0.59; in NWFP are 2.61 and 1.13; in Balochistan are 4.56 and 2.24 respectively. It
shows that people of low income groups in urban areas of Sindh have less consumption
of meat & fish as compared to other provinces.

169
 Rural Urban Food Consumption Analysis

7. Policy Implications
The results of the study indicate that the households of lower income groups in Pakistan
as well as in all provinces have higher expenditure elasticities of different food items as
compared to the higher income groups. It explains that the living standard of people in
the domain of lower income groups is very low. These households are unable to meet the
basic nutrients required for the family. For example, the consumption of meat & fish is
lower in the lower income groups in the rural NWFP households and urban Balochistan
households. Therefore, the government is suggested to launch such type of projects that
would raise the income level of these people and government may provide the necessary
food items to the poor families at cheaper rates.

REFERENCES

Aguiar, M., & Hurst, E. (2005). Consumption versus Expenditure. Journal of Political
Economy, 113(5), 919-948.
Cramer, J. S. (1957). A Dynamic Approach to the Theory of Consumer Demand. Review
of Economic Studies. 24. 73-86.
Fabiosa, F. J., &Soliman, I. (2008). Egypts Household Expenditure Pattern: Does it
alleviate a Food Crisis? Working Papers 2008:475, Center for Agriculture and Rural
Development, LowaStateUniversity.
Gujarati, D. (1998). Basic Econometrics. 3rd Edition, New York: McGraw-Hill
Hauthaker, H.S. (1960). Additive Preferences. Econometrica, 28 (3) 244-257
Houthakker, H. S. (1961). The Present State of Consumption Theory. Econometrica, 29
(4) 704-740.
Oguoma, N. N. O., Ehirim, N. C., Benchendo, C. N., Nnadi, I., &Okoronkwo, E. N.
(2010). Determinants of Aggregate Household Demand for Edible Oils in Imo State.
African Journal of Agricultural Research, 15(5), 1932-1936.
Obayelu, A. E., Okoruwa, V. O., & Oni, O. A. (2009). Analysis of rural and urban
households food consumption differential in the North-Central, Nigeria: A micro-
econometric approach. Journal of Development and Agricultural Economics, 2(1), 18-26.
Parpiev, Z., &Yusupov, K. (2011). Testing Household Economies of Scale in Uzbekistan.
Eurasian Journal of Business and Economics, 7(4), 25-51.
Pakistan, Government of (2001). Household Integrated Economic Survey 1998-99.
Federal Bureau of Statistics, Islamabad.
Syrovtka, P. (2007). Exponential model of the Engel curve: Application within the
income elasticity analysis of the Czech households demand for meat and meat products.
Agric .Econ - Czech, 53(9), 411420.
Stone, R. (1954). Linear Expenditure System and Demand Analysis: An Application to
the pattern of British Demand. Economics Journal, 64, 511-27.
Tey, Y. S., Mad, N, S., Zainalabidin, M., Jinap, S., & Abdul, G. R. (2009). Demand for
Quality vegetables in Malaysia. International Food Research Journal, 16, 313-327.

170