The analysis and examination of household consumption patterns made possible
by the pioneering work of Ernest Engel, in the form of the Engel curve,
1
is a critical
element for the formulation of various aspects of economic policy. This information is
essential for macroeconomic planning purposes, as accurate projections of demand for
various commodities are critical for efficient allocation of scarce productive resources
across the different sectors of the economy. Knowledge of consumer behavior is also
important for evaluating the impact of tax proposals on household welfare, as taxes
imposed on commodities having an income elasticity greater than one are likely to effect
rich households, while taxing necessities (with elasticity below one) will have a
disproportionately adverse effect on low income households.
A vast empirical literature has examined household consumption patterns, using the
Engel curve framework for both the developed and developing countries. Noteworthy studies
in this regard include Stigler (1954), Houthakker (1957), Giles and Hampton (1985) and
Tansel (1986). In case of Pakistan, household consumption patterns have been analysed by a
large number of studies, which includes Ranis (1961), Rahman (1963), Bussink (1970), Ali
(1981), Malik (1982), Cheema and Malik (1985), Malik and Ahmad (1985), Ahmed and
Ludhow (1987), Alderman (1988), Burney and Khan (1991, 1992) and more recently Shamim
and Ahmad (2007) and Ahmad and Arshad (2007). The major limitation of the existing
literature, apart from being based mainly on datasets which are over two decades old, is that
the household consumption patterns have been analysed only for Pakistan as a whole or by its
urban-rural regions. To our knowledge, no study has examined the consumption behavior of
households across the four provinces of the country.