Two-way fixed effects, without controls

	ratio of net migration to 1994 population
	(a)	(b)	(c)
SOE reform 1998*year 2002			-0.019
			(0.377)
SOE reform 1998*year 2001			-0.031
			(0.377)
SOE reform 1998*year 2000			0.005
			(0.377)
SOE reform 1998*year 1999			0.048
			(0.377)
SOE reform 1998*year 1998	2.810***		2.740***
	(0.324)		(0.377)
SOE reform 1998*year at and after 1998		0.608***
		(0.218)
SOE reform 1998*year 1997			-0.027
			(0.379)
SOE reform 1998*year 1996			-0.683*
			(0.377)
SOE reform 1998*year 1995			0.070
			(0.377)
Year fixed effect	yes	yes	yes
Prefecture fixed effect	yes	yes	yes
N	1,682	1,682	1,682
R2	0.048	0.005	0.051
Adjusted R2	-0.065	-0.113	-0.067
F Statistic	75.318*** (df = 1; 1503)	7.803*** (df = 1; 1503)	9.965*** (df = 8; 1496)
*p < .1; **p < .05; ***p < .01
This table shows the two-way fixed effect regression results of net migration rate on the interaction term between SOE refrom in 1998 and time dummies. All models include year and prefecture fixed effects. Column (a) shows that SOE reform in 1998 is positively associated with out-migration in the same year. Column (b) and (c) shows that the impact of SOE reform in 1998 on migration in and after 1999 is not statistically significant.