The validity and international comparability of China’s floating population concepts and data

The two concepts of “ liudong renkou (floating population or FP)” and “ renkou liudong (mobility of the floating population or MOFP)”, along with relevant data based on these two concepts, have long been used extensively in China’s research and policy making, playing a central role in Chinese studies of migration. Unlike the concepts of “migrant” and “migration” in the international literature, which are focused on people’s spatial mobility, “ liudong renkou ” and “ renkou liudong ” are identified and measured by the separation of one’s place of residence from one’s place of household registration ( hukou ), an approach inconsistent with relevant international practices. By analyzing various census data and data from the China Migrant Dynamic Survey (CMDS), this article examines the validity and reliability of these two concepts and the data based on them in the international context, revealing that they have become increasingly invalid and unreliable for the purpose of measuring migration events since China’s reform and opening up in the late 1970s. The results further demonstrate that these two concepts and the data based on them have become increasingly detached from real migration events and processes. They may become invalid by overestimating the volume of the mobile population, ineffective due to systematic omission of certain mobile populations (such as urban-urban migrants), or misleading as to the changing direction of migration flows. In addition, data on the floating population cannot be used to calculate migration rates and are not comparable in the international context. The concepts of “ liudong renkou ” and “ renkou liudong ” and data based on these two concepts may still need to be used in China for a long period of time due to the continuing existence of the hukou system and its roles in the provision of public services, social welfare and social security. However, we argue that concepts, measurements, and methods of data collection in research on migration in China should be gradually shifted to and focused on migrations as spatial events; further, transition data, based on an individual’s residence five years ago and one year ago, should be gradually adopted as the main data source and included in the short form of future censuses; additionally, migration event data based on population registration and administrative records should be used more fully, so that China’s migration research can be conducted on the solid basis of valid and reliable data sources.