Rapid economic growth and radical structural transformation pose a challenge to official
statisticians as they seek to encompass new economic activities and phenomena. The accuracy of
official statistics is liable to come into question. Urban unemployment in China is a good example.
The paper estimates the urban unemployment rate using administrative statistics, population
census data and a new sample survey data set, showing how and why Chinese unemployment
statistics are a minefield for the unwary and unemployment is so difficult to measure. Nevertheless,
it can be said that the urban unemployment rate exceeded 11 % in 1999 and 2000. The paper
concludes by considering the implications of the findings for understanding unemployment, for
policy, and for the collection of statistics.