There are three major problems with
urban-rural population data in China:
1. The Chinese household registration system (the so-called Hukou
system) is using a very peculiar and outdated classification system of
the population. Basically, everyone who is registered by a village
committee is classified as a farmer and belongs to the rural
population. Everyone, who is registered by a street committee in a
town or city, belongs to the urban population. Of course, this is
highly inaccurate because today there are many people still registered
in a village at their place of birth, who have in the meantime
migrated to a town or city and work in the non-agricultural sector.
This population with "false rural" registration might be in the range
of 15 million people.
2. On the other hand, there are people living in towns with an urban
household registration, who work in agriculture. They are essentially
farmers, who have moved to towns and suburbs and changed their
household registration from village to town - but they are essentially
rural people. This group with "false urban" registration might be in
the range of 23 million people.
3. There is also a very large group of people, who are registered in a
village (and thus are classified as rural) but work temporarily in
urban areas. This is the so-called "floating population" of rural
people, which can be easily observed in all major cities of China.
Experts have estimated this group of the population to be in the range
of 90 to 110 million. Perhaps some 65 million of this "floating
population" are true migrant workers, who spend some time in a big
city to earn money but still have a significant income from farming.
They regularly go back during harvest times. The rest of this migrant
workers are in different stages of actual migration. They more or less
permanently live in big cities and are "tolerated" despite their
village registration - such as many ten-thousand semi-permanent
migrants from rural areas in certain districts of Beijing.