Most of the empirical studies label the first term on the right hand side the ‘pure’ productivity effect, the second term as the ‘reallocation’ effect. [...] In the particular counterfactual chosen here, we assume that in the absence of the competition that leads to market-share change, both exiting plants and continuing plants would have remained in the market at the end of a period and that there would have been no changes in their output shares during the period. [...] This relative difference in the importance of turnover in the two components is also found over the two previous periods.9 Most of the within-plant productivity contribution comes from the plants that increased market share. [...] The greater the shifts in market share in a period, the higher was aggregate labour productivity growth in the period. [...] A slowdown in productivity growth in the 1979-1988 period relative to the 1973-1979 period occurred as the pace of shifts in market share 9 See Baldwin and Gu (2002) for an analysis of the components of entry that examine the difference between foreign and domestic plants, single enterprise and multi-enterprise plants.