Some issues in the measurement of comparative advantage

This paper examines measurement of 'revealed comparative advantage' in the light of recent evidence that such measures appear to yield poor results and of recent theoretical developments. Evidence is presented that the poor performance of these measures is not likely to be attributable to problems predominantly associated with measurement itself. Recent theoretical developments make clearer what can be expected of observed trade flows. A measure, based on theory's ex antepredictions measures. The behaviour of the measure is evaluated for a sample of UK industries in 1979.