The global food crisis of mid-2000s resulted in a several-fold increase in the prices
of essential food items. Resultantly, the incidence of food insecurity, hunger, and poverty
has increased in many developing countries [Ivanic and Martin (2008); Harttgen and
Klasen (2012); De Hoyos and Medvedev (2009); World Bank (2010); Regmi and Seale
(2010); Andreyeva, et al. (2010). Pakistan is also hit hard by this crisis. Prices of several
food items increased by more than a 100 percent since 2006-07. Consequently, nearly
half of the population is currently unable to meet its minimum (subsistence) caloric
requirements for healthy and productive living [Malik, et al. (2014)]. A large proportion
of household expenditure is spent on food (on average about 48 percent in 2010) and thus
very little is left for the other expenditures necessary for human welfare, such as, health
and education. Moreover, dietary diversity is extremely limited. Nearly 70 percent of
food expenditure is on cereals, dairy, sweeteners, and fats. Wheat is the major source of
calories, providing about half of the total daily calories [Malik, et al. (2014)]. However,
the price of wheat increased by 125 percent between 2005-6 and 2010-11. Existing
analyses indicate that these price shocks entail significant additional expenditures to
maintain their pre-crisis consumption levels [Haq, et al. (2008); Friedman, Hong, and
Xiaohui (2011)]. There is thus overwhelming evidence that rising food prices and the
decline in real wages have serious implications for poverty, food security, and nutrition
through food consumption patterns in the country.
In Pakistan, several studies have examined the effect of price change on
consumption patterns during the last four decades [Siddiqui (1982); Burney and Khan
(1991); Malik and Sarwar (1993); Burki (1997); Farooq, et al. (1999); Shamim and
Ahmad (2007); Haq, et al. (2008, 2011)]. However, the analysis in these studies is based
mostly on the data collected before the food price hike (i.e., before 2008). Some postprice-crisis studies, for example, Haq (2008, 2011) and Friedman, Hong and Xiaohui
(2011), provide useful information on the impact of food price crisis on the welfare of
Pakistan‘s population. However, these studies are limited in several ways by the
assumptions underlying their analysis. For example, they assume similar consumption
patterns across different household expenditure groups and across different regions of the
country; and, thus fail to highlight the differential impact if any of the food price hike on
the consumption patterns of poor and non-poor households located in different regions of
the country. A fuller understanding of the consumer response to rising prices based on
disaggregated analysis is essential for the policymakers to design effective and pro-poor
food policy in the current scenario.
The main objective of this paper is to examine the extent of the impact of more
recent price changes on consumer behaviour at a disaggregated level and highlight the
policy implications for poverty, food security, and nutrition in Pakistan. For this purpose,
using the data of the most recent publicly available and nationally representative
Household Integrated Economic Survey (HIES) 2010-11, we estimate the Linear
Approximate Almost Ideal Demand System (LA-AIDS) for ten food groups: wheat and
wheat flour; rice including all kinds of rice consumed; other cereals; pulses; fruits and
vegetables; milk and milk products including desi ghee and butter; meat (beef, mutton,
fish and poultry); edible oil; sugar and other sweetener; and other food items (tea,
condiments and spices, etc.). We divide households into two groups: poor and non-poor,
and differentiate for rural and urban areas.
This paper is divided into five sections. Methodology and data are described in
Section 2. A descriptive analysis of food consumption patterns is presented in Section 3.
Section 4 presents a discussion of the results of LA-AIDS model and estimated
elasticties. Implications of food consumption patterns for poverty reduction are presented
in Section 5. Conclusions and policy recommendations are given in the final section.