This study departs from the previous literature on purchasing power parity (PPP) by proposing a demand
system based methodology for calculating the PPP that takes account of consumer preferences and allows for
the substitution effect of price changes. The methodology is applied to provide evidence on PPP between the
Indian Rupee and the Vietnamese Dong. The study is conducted within a framework that allows for regional
variation in preferences and price changes both inside the country and between countries and proposes and
applies a methodology for constructing prices from unit values after adjusting them for quality and
demographic effects. Using these prices the intra-country PPPs for India and Vietnam are calculated using
the single equation (Engel curve based) procedure of Coondoo, Majumder and Chattopadhyay (2011). The
cross country PPPs are calculated between sectors and across expenditure classes, apart from PPP at
aggregate country to country level, using both the single equation and system based procedures. The paper
contains evidence that the incorporation of price effects leads to a significant change in the PPP rates
obtained from using cross section data (single equation procedure) ignoring price changes. The demand
system based methodology yields PPP rates that are consistent with those obtained from conventional
procedures such as the CPD method, yields standard errors of the PPPs and has the additional advantage of
testing for invariance of inter-country PPP across expenditure classes. The disaggregated PPP rates question
the conventional practice of using a single economy wide PPP in inequality and poverty comparisons.