摘要

This article examines the last (rightmost) digit of China's prefectural GDP growth rates, and find that they are unevenly distributedamong 0 to 9. There are too many 0s, 1s, 2s, 5s and too few 4s, 7s, 9s. Moreover, the distribution of the last digit is less skewed when the macro economy grows quickly and when the prefecture's growth rate ranks high within the province. These patterns reveal that local officials artificially influence the data in order to exceed some thresholds. However, they do not necessarily imply that China's GDP data are falsified.