In the study of terminations, we need to make sure that at least some of the terminations are disciplinary terminations rather than normal retirements, in particular for the post-1985 sample. Comparing the whole sample with the post-1985 subsample (Table 2), we can see that forced retirement at the age of 65 indeed has a significant role. No provincial leaders were older than 68 in the post-1985 subsample. Moreover, in the post-1985 subsample, a significant number of terminations happened at exactly the age of 65 (25% of the termination cases). However, the age of retirement rule was still not strictly enforced. Even in the post-1985 subsample, for the group of leaders at age 65 or above, there are more cases of leaders remaining at the same level (24 cases) than terminations (21 cases). Two leaders were even promoted at the age of 66. Examining the termination cases only, a significant number of terminations (more than 40% of the total number) occurred beyond the retirement age of 65. The fact that some provincial leaders were terminated at the “normal” retirement age or even younger, but others stayed or were promoted at the same or older ages implies that certain variables other than age must have played a disciplinary role in leader turnover. We will examine whether economic performance is such a disciplinary variable in Section 5.
Table 1 also shows that, in the period 1979–1995, China achieved a very rapid growth rate. The average annual growth rate was more than 10%.18 However, the growth rate varies widely across provinces and over time with a standard deviation of 6%. The fastest growth rate (more than 50%) occurred in Yunnan Province in 1994, while the slowest (negative 16%) occurred in Gansu Province, also in 1994. The large variation of the growth rate is important for us to identify how it is correlated with leader turnover.
Other independent variables are also described in Table 1. More than half (62%) of provincial leaders have a college degree, which shows that the central government indeed tried hard to promote well-educated cadres. About 23% of provincial leaders have connections with the center. The average tenure of office in the sample is 3.03 years, with some leaders staying as long as 12 years. Since the official term for a provincial leader is five, the average of 3.03 years indicates a considerable degree of turnover