This paper uses nationally representative household income and expenditure survey data for Pakistan to
investigate how the receipt of international remittances affect the average and marginal spending behaviour of
households on five different categories of goods: food, education, health, non-durables and durables. Two
findings emerge. First, expenditure share on food for households that receive remittances would have been more
if the households had not been receiving remittances. Similarly, less spending on the other four categories of
education, health, non-durables and durables is predicted for remittances- receiving households had they not
been receiving remittances. Second, households that receive remittances spend less at the margin on food and
durables and more on education, health and non-durables. At the mean, compared to households that do not
receive remittances, the households receiving remittances spend, at the margin, 10 per cent and 4 per cent less
on consumption of food and durables, respectively. Moreover, the marginal increase in spending on education is
26 per cent more for a remittances-receiving household than for a non-receiving household. Finally, the
households receiving remittances spend, at the margin, 14 per cent more on non-durables (which includes their
spending on housing, and is thus akin to investment in physical capital) than the households with no
remittances. A key feature of these results is the likely positive impact of remittances on economic development,
by the way of increased spending on human capital or education as well as physical capital. Remittancesreceiving
households appear to look at the remittance earnings as a transitory income and therefore tend to
spend remittances more on investment than consumption. This finding lends support to the permanent income
hypothesis.