Variations among SOEs
The statistics reported in table 7.3 show a limited degree of variation in industrial SOE attrition rate among different provinces, though the coefficients of variation were relatively higher during the early years of restructuring. There was a similar trend of declining variability among different prefectures and among cities and counties. But the degrees of variation were greater at these levels of comparison, especially at the outset of the restructuring. What explains such a pattern of variability? An analysis of a large-scale national survey of ownership restructuring conducted by the National Bureau of Statistics among industrial SOEs (n = 40,238) in 1998 reveals that fast movers in the early phase of restructuring were those that had lighter financial and personnel liabilities and stronger profitability than laggards and nonstarters (Lin and Zhu 2001). What the finding suggests is that, although the decision on ownership restructuring was largely motivated by the problems of poorly performing SOEs, the pace of change did not bear a positive correlation with the seriousness of the problems they faced. Rather, many less unhealthy enterprises were actually rushed out of the door first. A likely driving force behind this is a political bandwagon response, highlighted above, of local officials to the centrally adopted policy change, though their abilities to respond were unevenly constrained by the varying restructuring costs associated with the preexisting organizational conditions of the enterprises under their purview.
For local governments facing similar constraints, on the other hand, their strategies toward privatization were subject to peer influence. As political rivals, local officials sought to outshine each other and would rarely turn a blind eye on the policies of their regional peers. The more risk-averse actors also tended to follow the footsteps of those who were politically more entrepreneurial, especially when the latter’s deviant behavior (e.g., not focusing on mergers within the public sector) not only went unpunished but reaped benefits. In the 2004 CASS-HKUST survey of private enterprise owners (n = 511) mentioned above, 55% of the respondents agreed or strongly agreed that the changing attitude of their local governments toward private ownership in the preceding decade was in part due to some kind of demonstration effect from the more liberal policies adopted by governments in neighboring regions. While such an effect would hasten the tempo of privatization, it may also hold a clue to explaining the convergence that quickly followed the initial disparity in the decline of SOEs.
To examine this possibility I analyze the data from NBS annual industrial surveys during the ownership restructuring process in 1998-2005. The focal issue is how regional competition and learning may have affected the attrition rate of industrial SOEs, which is defined as the ratio of the current- year number of industrial SOEs in a locale to that of the preceding year. The
measure I use to proxy regional peer influence in post-1997 privatization is the local attrition rate of industrial SOEs divided by the regional attrition rate, both lagged for one year. A ratio less than 1 implies a lagging position and thus pressure for catch-up, whereas a ratio greater than 1 implies a leading position and thus a gap to be narrowed by laggards. A number of control variables are included in the analysis. Descriptions of these variables, along with regression method and results, are posted at the book site.
The data analysis yields three basic findings. First, the pace of ownership restructuring of a locale relative to the regional average in a given year was negatively correlated with the local attrition rate of industrial SOEs in the subsequent year. What this suggests is that many laggards probably played catch-up, whereas the lead of faster movers tended to be short-lived. Second, such effect was similarly significant among cities and counties within the same prefecture and among prefectures in the same province. The regional influence on local attrition rates thus appears to have been widespread and contagious. Third, the significance of the effect of the explanatory variable weakened over time. This reflects a growing convergence, resulting from earlier regional peer influence, in the attrition rate of industrial SOEs among different locales. A common outcome of these forces is that they pushed ownership restructuring far beyond the initial plan of the central leadership.
-  An enabling condition for this was furnished by the concurrent intensification of the efforts of government authorities and SOEs to find alternative placement for existing employees, including relaxingrestrictions on private business. From 1998 to 2002 a total of 27.2 million state sector employees werelaid off through various furlough arrangements that eventually channeled them out of the urban publicsector (LSYC 2003, 134).