Desktop version

Home arrow Political science arrow Dancing with the devil : the political economy of privatization in China

Variations among TVEs

Given the important role of TVEs in the public sector before 1997, it is equally interesting to further examine what mechanisms were in play among these enterprises during their post-1996 acceleration of privatization. The change there was more precipitous than that among SOEs, as noted at the beginning of this chapter. Chapter 5 shows that privatization had actually started earlier among TVEs, though the pace was constrained by the political risks involved and varied according to the different abilities of local authorities to justify such a move with demonstrable results from the expansion of the space for private business. Such political constraint was lifted by the policy change at the Fifteenth Party Congress.[1] Although TVEs were not explicitly mentioned as part of the plan for ownership restructuring, many local authorities sensed the new direction in central policy and pressed ahead with their owngaizhi programs (Zhu Rongji 2011, 2:501-502). The Ministry of Agriculture (MOA), which was the central policymaker for TVEs, also moved in sync and gave the green light for such undertaking.[2]

From 1997 to 2003 the average annual attrition rate of TVEs was 21% (CTEY, various years), resulting in the reduction of the number of TVEs by 1 million. In this process of change TVEs were prone to the same kind of regional peer influence faced by SOEs. But unlike many SOEs that encountered difficulties in severing organizational ties with their employees (and thus were unable to seize upon the opportunity of restructuring immediately), the privatization of TVEs was not constrained by old labor institutions inherited from the Mao era. As Walder (1995) has noted, from the very beginning of reform employment relations in TVEs were largely driven by labor market conditions and not bound by the kind of social service and welfare responsibilities appended to urban public enterprises. As a result, the initial tempo of privatization among TVEs following the party congress tended to be more heavily influenced by the preexisting strength and mix of public and private ownership in the local nonfarm sector.

In chapters 4-5 I have outlined three different local patterns of ownership structure and change: (a) locales where the sales growth strategy was forcefully carried forward up to the time of central policy change and public enterprises dominated the nonfarm sector; (b) locales where the strategy had stalled but the private sector remained underdeveloped despite the languishing of the local public sector; and (c) locales where the initial dominance of the strategy was eclipsed by subsequent growth of private enterprises. For local officials in pattern (a), the policy change in 1997 opened up an opportunity to unload poorly performing public enterprises. But they would probably not have very strong incentive to immediately privatize those that were still able to maintain the existing modus operandi and thereby help address the revenue and employment imperatives. For local officials in pattern (b), the reduced political risk for privatization created a major impetus for quick action to let go of local public enterprises, most of which had been kept for lack of viable alternatives yet gradually lapsed into increasingly unbearable liabilities for the local governments due to sagging performance. For local officials in pattern (c), where many poorly performing public enterprises were likely to have already been rid of before 1997, the rational choice was to trim those that had been retained as window dressing while keeping those that had been and could still be of some use for various reasons. It follows that the initial response of local officials to the central policy changes was likely to be relatively modest among the strongholds of the erstwhile dominant strategy and among steady early movers of privatization, but it was likely to be very strong among laggards in both regards. Yet, once the effect of regional peer influence kicked in and intensified, such initial difference tended to fade.

A case in point for investigating these possibilities is Jiangsu. It was famed for being an exemplary model of TVE-driven economic growth up until the mid-1990s (Ho 1994; see also Whiting 2000). It is also a province for which systematic data at the county level can be found to trace both the changes among TVEs and their operating environments in the first two decades of reform. Like those in other provinces, TVEs in Jiangsu eventually succumbed to the changing policy and economic conditions around the turn of the century. Their pattern of decline thus helps illuminate some important mechanisms at work.

Figure 7.3 illustrates the rise and decline of TVEs in Jiangsu. The decline started in the early 1990s. It began to accelerate after 1995 and became precipitous after 1997. Table 7.4 further suggests that many of the enterprises that led the way down were those that had experienced substantial expansion in earlier years, as may be inferred from a contrast of the numbers for larger enterprises (with annual sales above 5 million yuan) with those for industrial TVEs of all sizes in 1995 and 1998. Table 7.4 also shows that, while the overwhelming majority of TVEs disappeared in all three regions of the province by 2004, the initial decline was much slower in southern Jiangsu, the heartland of TVE-led growth in the 1980s and early 1990s. In contrast, the rush to privatization immediately following the policy change in 1997 was faster in central Jiangsu and much faster in northern Jiangsu, where some locales had experienced significant slowdowns of TVE growth and/or even a major expansion of the private sector earlier in the decade. This is consistent with part of what is postulated above. Yet there were nevertheless considerable

Total number (1,000s) of TVEs in Jiangsu, 1984-2004 Sources

FIGURE 7.3 Total number (1,000s) of TVEs in Jiangsu, 1984-2004 Sources: JSRSY2002; CTEY2003,2004,2005.

variations among different prefectural cities within all three regions in the initial attrition rate of TVEs, as can be seen from the prefectural figures in the table.

To investigate these variations and the subsequent convergence in the outcome of decline among TVEs in the province, I probe below the prefectural level and analyze county-level data, which provide more revealing information about the sources of the forces at work. The main questions are what accounted for the different attrition rates in the initial push for privatization and how regional peer influence may have affected the pace of decline over time.

For the analysis of the first issue I run a pooled regression spanning the period of 1998-2000. I use the attrition rate of all industrial TVEs in each county-level unit of the province as the dependent variable. The main explanatory variable is the initial policy stance of each unit toward privatization in terms of the three major possibilities discussed above: stronghold of TVE- driven growth, early mover in privatization, and laggard in both regards. The hypothesis under testing is that laggards tended to move faster than the other two types of locales in the immediate aftermath of the central policy change. For the regression analysis on the issue of regional peer influence, I use a panel data set that spans the period of 1998-2004. The basic setup is similar to that for the analysis of regional peer influence among industrial SOEs in the preceding section. The dependent variable is the attrition rate of TVEs, and the main explanatory variable is the ratio of county-level attrition rate to

Table 7.4 Decline of TVEs in Jiangsu province

Regional unit

Number of TVEs with annual sales above 5 million yuan

Average annual attrition rate (%) in 1997-1999





19,776 (93,476)

3439 (64,468)

1,039 (3,994)




8,524 (44,235)

1,630 (35,762)





2,289 (12,091)

392 (9,090)




2,066 (13,620)

383 (9,807)




1,623 (7,667)

295 (8,367)




1,269 (4,913)

252 (4,043)




1,277 (6,144)

308 (4,455)




5,838 (21,373)

739 (14,244)





2,890 (13,731)

374 (4,141)




23i (4,579)




2,948 (7,642)

134 (5,524)




5,414 (27,668)

770 (14,562)





1,602 (8,441)

205 (4,676)




1,981 (6,827)

88 (1,130)





127 (i,708)




959 (7,38:)

116 (2,948)




872 (5,019)

234 (4,100)



Notes: (a) The figures in parentheses are for all industrial TVEs in 1995 and 1998, and the 2004 provincial figure (italicized) in parentheses is for township-owned (including nonindustrial) enterprises of all sizes; (b) annual attrition rate during 1997-1999 is for all industrial TVEs; (c) Huaian prefecture was known as Huaiyin before 2001. Taizhou and Suqian were not prefectural cities before 1996 and belonged to Yangzhou and Huaiyin respectively.

Sources: YJSTE1997-2000; CTEY2005; 1998 NBS annual industrial survey data; 2004 economic census data.

the prefectural average rate in the preceding year. Descriptions of the other variables included in this analysis and the preceding one, along with discussion of regression methods and summary of the main estimation results, are posted at the book site.

What the pooled regression reveals is a significantly positive effect of the leading role played by laggard counties in the initial rush toward privatization. This is consistent with the conjecture developed earlier. There is also evidence from the panel data analysis that confirms both the effect of regional peer influence on TVE attrition rates and the weakening of such effect over time. What these findings illustrate is that the initial responses of TVEs to the central policy change in 1997 were patterned by the lasting impact of their recent trajectories of development under different mixes of ownership. Like SOEs, however, the fall of TVEs thereafter was not simply a locale-specific affair, but rather a process where the pace of change followed interdependent calculus used by officials across different locales within a larger ecosystem of political governance. In tandem these two elements (initial leading role of “laggards” and growing regional peer influence) of the political bandwagon reactions among TVEs added to the precipitousness and the breadth of the decline of public ownership beyond the limits set by the central leadership in 1997.

  • [1] An indication of the reduced political risk for privatization in rural areas is the diminishing relevance of shareholding cooperatives, which had been widely used as a way to disguise privatization ofTVEs before the policy change at the Fifteenth Party Congress (Tsai 2002; Whiting 2000). In 1996there were 1.55 million TVEs and 0.144 million rural shareholding cooperatives. The number of the latter went up to 0.194 million in 1997 but quickly turned south along with the decline of TVEs. By 2002the total number of TVEs dropped to 0.401 million, whereas that of rural shareholding cooperativeswent down to fewer than 80,000 (CTEY1997, 129; CTEY 1998, 174; CTEY 2003, 186).
  • [2] The MOA’s new policy guidelines and directives on ownership restructuring among TVEs can befound in the 1999-2002 issues of the China Township Enterprise Yearbook.
< Prev   CONTENTS   Source   Next >

Related topics