Date of Award

Winter 1990

Project Type


Program or Major


Degree Name

Doctor of Philosophy

First Advisor

Evengelos E Simos


For nearly a century, economists have debated the choice of appropriate prices in the empirical examination of the purchasing power parity (PPP) of exchange rate determination. The conventional price indices suffer from bias in measurements caused by the varying aggregation methods employed in different countries. The absence of uniform price indices across countries tends to frustrate empirical probing of the purchasing power parity theory. The problem is exacerbated by including non-traded goods in the composition of the conventional price indices used to test the PPP.

In this thesis, the problem of the 'index-numbers' is overcome by creating a unique set of price indices. These price indices are based on the purchasing power parity conversion factors obtained from the outcome of a joint project undertaken by the Organization for Economic Cooperation and Development (OECD) and the European Statistical Office (EUROSTAT) in 1985.

Using the total economy (aggregate) and the tradable-goods versus nontradable goods (sectoral) prices, the conventional and the proposed modified purchasing power parity-based models of exchange rate determination are tested. The data are quarterly and for most of the currencies tested, they range from the first quarter of 1973 to the second quarter of 1988. Five bilateral exchange rates, four involving the Dollar rates and the one the Pound rate are empirically examined.

Because of the endogeniety of prices and exchange rate, the instrumental variables technique is applied to all models. The results are supportive of the sectoral approach and the tradable-goods hypothesis formulation of the purchasing power parity theory. To investigate the effect of structural change on the exchange rate behavior, the basic purchasing power parity models are modified to include the relative internal price ratio. The results support the notion that, in the long-run, purchasing power parity is essentially a monetary phenomenon. As expected, internal price variations caused by real economic fluctuation do not alter the outcome of the models tested.