On May 11, the National Bureau of Statistics (NBS) released the preliminary results of China’s 2020 National Population Census. In addition to confirming what demographers have long suspected — China’s population is aging, urbanizing, and the ratio of boys to girls at birth is evening out — one result stood out: China’s “floating population” of internal migrants, once believed to be on the decline, is still growing rapidly.

More specifically, the census found 493 million people in China living outside of the townships or streets where they are officially registered. That’s roughly one-third of the entire population, and 55 million more than in the 2010 census. Of these, approximately 117 million people are living in different parts of the cities in which they are registered, while the remaining 376 million have migrated even farther afield, including more than 124 million who have left their home provinces or province-level regions altogether.

The first thing this tells us is that the proportion of people migrating from rural to urban areas is not falling, but rising.

Second, the data suggests more migrants are opting for short-distance migrations. According to the 2020 census, there are around 251 million people who migrated within their home provinces, nearly twice that in 2010. The number of those who migrated within the same city tripled during the past decade.

In other words, China’s population is still on the move, even if the patterns have changed. But given that migration has defined the past 40 years of Chinese history, why should that come as a surprise?

To begin with, the census data contradicts the results of other, less representative surveys conducted over the past decade, suggesting that these interim surveys may have misjudged migration trends — and potentially misled policymakers.

Since 1990, in addition to conducting a population census every 10 years, China performs what it calls a “1% National Population Sample Survey” at the midway point between each census. Compared with the 2010 census, the figures from the 2015 sample survey showed only a slight increase in the floating population, suggesting that migration had reached an inflection point. Building off this, scholars and bureaucrats found in followup surveys that the migrant population of China’s eastern coast was beginning to decline after decades of explosive growth as migrant workers “returned” to their hometowns or opted to migrate closer to home.

The fluctuations in China’s recorded floating population between interim surveys — conducted by the NBS and the National Health Commission (NHC) — and the official census conducted last year suggest the need for further analysis to figure out what, if anything, went wrong. In the meantime, it seems that the 2015 sample survey, as well as the reports from the NBS and the data from NHC, cannot be used to accurately judge changes in the floating population over the past 10 years.

The purpose of censuses and population surveys is to provide statistical information to support the formulation of social and economic policy. When the underlying data is inaccurate, the policies built atop that data will not have their intended effect. The miscalculations from the previous five years are bound to affect the provision of public services to urban residents over the next five years, and only quick action from the relevant bodies can address the problem and help destination cities cope with the unexpected growth of their floating populations.

In 2012, scholars used data from the 2000 and 2010 censuses to argue that migration patterns in China were in line with the “Matthew effect,” which can be summarized as the rich getting richer. Wealthy coastal provinces were drawing residents from the northeastern Rust Belt and less-developed interior, and as these regions lost human and economic capital, the coast only grew more attractive. Data from the 2015 sample survey, on the other hand, suggested that migration trends could be on the verge of reversing, with the east seeing a decline in migrant numbers and the central and western regions showing an increase. The so-far limited data we have from the 2020 census shows that the 2015 sample survey was off, and the earlier trend toward population agglomeration along the developed eastern coast has remained steady over the previous decade.

In particular, the census found 10.9 million more people living in the southern Guangdong province than was predicted by a 2019 sample survey, likely due to an under-sampling of the migrant population. That same survey underestimated the populations of the eastern Zhejiang and Jiangsu provinces by a combined 10.12 million. Any education or other policies based on the original flawed data sets — whether in destination provinces or areas where migrants are leaving — are therefore hopelessly out of line with actual demand, and should be adjusted as soon as possible.

Steps should also be taken to address the provision of public services long term. China currently allocates resources primarily based on household registrations and residence permits. But even if people settle in a particular city, they do not necessarily live in the town, village, or subdistrict where they are registered. According to the 2020 census, China has 117 million urban residents living in their home cities but at different residences from what’s listed in their household registration files, not to mention the 376 million people who have migrated to other cities altogether. This shows how difficult it is to solve the real needs of the floating population based on household registration data, and should push the government to rethink its allocation strategy.

Finally, although China’s floating population is still largely made up of rural-to-urban migrants, there has been an uptick in the number of migrants between cities, and this migration is driven by a very different mechanism than its rural-to-urban counterpart.

China’s dual urban-rural structure means that social welfare and public services are less developed in the countryside than in the cities, so farmers with little in the way of specialized skills outside of agriculture are generally willing to move to whatever city will have them. However, after years spent working in cities, migrants’ level of specialization gradually increases, making it more difficult to match workers and employers. The larger a city is, the easier it is to find a match, and the greater the benefits of individual specialization.

The 2020 census data indicates that China isn’t an exception to the Matthew effect of population growth begetting more population growth. If anything, this trend may become more visible in the labor market over the coming years. This will increase the competitive pressure on mid-sized cities, as they can either expand their populations to increase their competitiveness in the labor market, or else watch their populations fall and risk falling behind in China’s urbanization and modernization drive.

This article was co-authored by Pan Zehan, an associate professor at the Institute of Population Research, Fudan University.

Translator: David Ball; editors: Wu Haiyun and Kilian O’Donnell.

(Header image: Two workers at a construction site in Beijing, May 8, 2021. Greg Baker/AFP/People Visual)