The familiar 'Dollar Index' of country 'openness' is obtained essentially b
y attempting to remove the influence of nontradables prices from data on co
untry price levels, with residual price differences hypothesised to measure
trade impediments. Differences in nontradables prices are in turn hypothes
ised to reflect differences in factor endowments, which Dollar proxies by G
DP per capita. This paper examines both the theoretical and empirical biase
s introduced to estimates of nontradables prices by using this proxy. We sh
ow that, while the approach provides a reasonable approximation on average,
relatively agricultural-land or labour abundant countries are particularly
poorly captured by GDP per capita. (C) 1999 Elsevier Science B.V. All righ
ts reserved.