SOE reform in 1998 and migration in China

Abstract:

This study explores the causal relationship between SOE reforms and migration in China. It found that central policy to downsize SOEs in 1998 increased out-migration from prefectures that witnessed large declines of SOE employment. Controlling for GDP per capita, share of non-state sector employment and inward FDI does not change the results. Moreover, increase in out-migration took place promptly after SOE downsizing, but also died down quickly. SOE employment decline in 1998 resulted in surge of out-migration in the same year, but from 1999 on the impact of SOE downsizing became indistinguishable from zero. This suggests that many SOE workers who were skilled enough to out-migrate and find jobs elsewhere were probably waiting for national policy of state sector downsizing well before 1998 and grabbed the opportunity to exit their work units smoothly. On the other hand, it is rare for people who failed to out-migrate right after being laid off to find jobs elsewhere years later.

This study argues that improving labor market liquidity and reducing costs of migration are key to promoting re-employment of redundant SOE workers in the non-state sector. Firstly, exit mechanisms should be established in over-staffed state firms to allow workers who can find higher paid jobs elsewhere easily quit. Secondly, the hukou regime should further be loosened to establish a nationally unified labor welfare system, which would make it easier for people to migrate without worrying about losing access to social security and public services. Moreover, the rural-urban divide in the hukou regime should be abolished to protect rural migrant workers from being exploited and provide laid off SOE workers a level field of competition in the labor market.

Key visualizations

average net migration rate versus SOE downsizing in 1998, by half

average net migration rate versus SOE downsizing in 1998, by quartile

SOE downsizing in 1998 as proportion to 1994 population

net migration in 1998 as proportion to 1994 population