THREE
A Great Reversal
I used two phrases to describe rural reforms in the 1980s of the last century: "Advancing of non-governmental sector and retreat of the state." As for the 1990s of the last century, I also have two descriptions - "Advancing of the state and retreat of the non-governmental sector" and "retreat of rights of the people and advancing of the rights of the government." In the 1980s, the standard of living of the peasants improved day by day and the level of tensions was low in the rural area. In the 1990s, although rural economy continued to develop, the livelihood of the peasants was difficult and the level of tensions in the rural area accelerated considerably.
- Li Changping (2005), a former Party secretary in Jianli county of Hubei province
We began the last chapter with the story of Mr. Nian, the impoverished and self-deprecating rural entrepreneur who created a marketing storm in the early 1980s with his Idiot's Seeds. The decade also ended with him. In September 1989, Mr. Nian was arrested. The local procuratorate ofWuhu city where Mr. Nian's business was headquartered charged him with crimes of corruption and "embezzlement of state property." In 1990, the city government shut down his firm. It was an inglorious end to a once sensational brand - and brand-name - of Chinese indigenous capitalism.
The charges against Mr. Nian were so trumped up that they did not even pass muster with the low evidentiary threshold of the Chinese courts. The municipal court of Wuhu city found Mr. Nian guilty on the corruption and embezzlement charges, but the intermediate court of Anhui province overturned the verdict. Mr. Nian, after all, was running a private firm so the charge of "embezzling state property" was rather strange in the first place. The intermediate court found him guilty on charges of something else - hooliganism - and sentenced him to three years in prison for having had immoral relationships with 10 women between 1984 and 1989. (Upon hearing the verdict, Mr. Nian reportedly retorted, "No, twelve.")
109
110
Capitalism with Chinese Characteristics
The fate of Mr. Nian signified a new economic and political era for China's rural entrepreneurs. Politically, the post-Tiananmen conservative leadership mounted a nationwide crackdown on China's private sector. This eased somewhat in 1993 and 1994 and substantially thereafter, but a more binding constraint began to impinge on rural entrepreneurship for the rest of the 1990s (and still persists to this day). The prevailing economic policy in the 1990s was to favor the urban areas over the rural areas and to favor foreign capitalists - FDI - over indigenous capitalists. There are other components of this policy model as well, including an emphasis on technocratic development and an industrial policy approach in favor of large firms. The cumulative effect of all these policies was a dramatic change in the balance of power between the two Chinas - the rural China that is more capitalistic and market-driven and the urban China that is more state-controlled. In the 1990s, the balance tilted decisively in favor of the urban China.
Fittingly, 1989 marked both the decadal and policy turning points. As is well known, 1989 was a year of political turmoil. Young university students protested and held hunger strikes on Beijing's Tiananmen Square for almost two months before the authorities decided to forcibly clear the square. In this book, my focus is on the economic implications of what famously is known as the "Tiananmen crackdown." Most Western analysts, although acknowledging the political importance of the Tiananmen crackdown, view the event as a brief pause in China's economic reforms. Barry Naughton, in his textbook on the Chinese economy, has a section devoted to this period entitled, "The Tiananmen Interlude." He notes (2007): "The conservative attempts to roll back reforms were completely without success, however, and are often forgotten."
Naughton is correct that the immediate ideological effect of the Tiananmen crackdown was short-lived. The political assault on the private sector fizzled out fairly quickly after Deng Xiaoping's famous "Southern Tour" in 1992. But, the quote by Li Changping - noted at the beginning of this chapter - that reforms in the 1990s stagnated or even reversed themselves reveals a different judgment, at least about rural China. Li Changping is intimately familiar with rural China. He began his career as a rural official in 1983 and, by the late 1990s, he was the Party secretary of a rural county in Hubei province. We should take his view seriously.
The issue is whether Li's judgment can be supported by aggregate data. It is not sufficient to cite the view of one official - even one as knowledgeable about rural affairs as Li Changping. It is possible that Li was simply noting one data point about what was happening in his own county, and we should
A Great Reversal
111
not extrapolate the observations of a region to the entire country. As in the rest of this book, I have gone through pages and pages of government documents, including more than 20 volumes of detailed bank documents, directives, and regulations that as far as I know, have never been analyzed by Western academics. I have also assembled a dataset that consists of household income surveys and private-sector firm surveys. The totality of both the qualitative and quantitative evidence strongly supports Li's judgment that there was a substantial policy reversal in the 1990s, primarily in the rural areas.
My conjecture, although not yet an empirically settled view, is that the economic effects of Tiananmen were actually far more long-lasting than we previously believed. One observable effect of Tiananmen is that the backgrounds and outlooks of Chinese leaders before and after Tiananmen were very different. Many of the reformist leaders before Tiananmen - symbolized by Zhao Ziyang in particular - had first gained national prominence as a result of their economic stewardship of the poor rural provinces. More than any other politician, Zhao Ziyang and Wan Li had launched China's transformative rural reforms, in Sichuan and Anhui provinces, respectively. The leaders of the 1980s represented the rural and, by implication, the more market-driven part of China. Their demise was behind the increasingly urban biases in the orientation of China's economic policies in the 1990s.
The leaders of the 1990s came from entirely different backgrounds. The two top leaders, Jiang Zemin and Zhu Rongji, had built their political credentials in the most state-controlled and least-reformed urban bastion of China - Shanghai. They were consummate technocrats, trained as engineers, and they had spent many years primarily in the SOE system. Shanghai, as I detail in the next chapter, represents the apex of the urban bias in a political and economic system already laden with urban biases. Their ascendancy to national positions was a direct result of the political aftermath of Tiananmen.
The second policy development after Tiananmen was a significant reversal of the policy approaches prevailing in the 1980s. The sharpest reversal occurred in rural finance. Many of the very productive financial experiments were terminated and credit constraints on small-scale, low-tech, and labor-intensive rural entrepreneurship were tightened. There were other policy reversals as well, including backtracking on political reforms and centralizing the administrative and fiscal affairs of rural governance, and a fiscal recentralization in 1994 that substantially reduced the autonomy of the provinces. (The latter development is under-researched. A much-heralded Chinese innovation, "federalism, Chinese style," in fact ended in
112
Capitalism with Chinese Characteristics
1994, with potential long-term harmful effects.1) The cumulative effects of all these policies brought a change in the trajectory of capitalism in China. In the 1980s, a version of market-driven, small-scale, and welfare-improving rural capitalism was developing vibrantly. In the 1990s, China continued to march toward capitalism but toward a very different kind of capitalism. Whereas Chinese capitalism in the 1980s was a rags-to-riches capitalism, the capitalism in the 1990s led to sharp income inequalities, a reduction of social opportunities available to the rural population, slower income growth, and an investment-heavy growth pattern. In this chapter, I begin by describing the sharp reversal in the fortunes of the hundreds of millions of rural Chinese in the 1990s. In the 1990s, their income growth slowed down considerably compared with the 1980s. One notable development was that rural entrepreneurship experienced many difficulties. While farming became increasingly less lucrative, rural Chinese flocked to labor markets to contribute their labor rather than to start their own businesses. This was so despite the fact that the returns from entrepreneurship exceeded labor income and that labor income, adjusted for inflation and time allocation, in fact declined drastically in the 1990s. The TVEs began to languish after the mid-1990s, and they languished as essentially private firms rather than as collective firms.
Sections two and three of this chapter describe the policy and political dynamics behind this monumental change in China's countryside. I first contrast developments in the financial sector between the two decades. A little-known fact about the 1980s is that significant and imaginative innovations occurred in rural finance. The pro-private sector tilt in policies regarding access to bank credits was substantial. There was also an explicit effort to permit private entry into the financial sector and to benchmark the performance of the state-owned banks against that of the informal financial sector. All of this was to change in the 1990s. Decision making in loan policies became more centralized and many of the productive financial innovations were discontinued. The state reversed its stance on informal finance completely, no longer viewing private credits as a useful complement to formal finance but rather as lethal competition to be firmly stamped out.
Section three of this chapter describes other broad changes between the two decades. In particular, the promising political reforms that began in the 1980s were thoroughly repudiated. Those reforms had aimed at fostering some intra-Party democracy and moving in the direction of more accountability of the CCP. Between 1989 and 2002, these reforms stagnated completely. There was a massive effort to recentralize the fiscal and political management of China's vast countryside. Last, but not least, an important
A Great Reversal
113
development in the 1990s was the rise of a technocratic state. A core feature of that state was that it viewed economic development in engineering terms and that it had the strength of conviction - and invested a huge amount of political capital - to force through its policy visions, at staggering immediate political and social costs and, in the long run, at huge economic costs as well. Above all, this was a state that went to extraordinary lengths to empower itself both politically and economically. All of these developments reinforced the effects of the increasing financing repression of the rural private sector and were directly responsible for the social and economic malaise in the form of what came to be known in China as the "three rural crises" -the crisis of agriculture, the crisis of rural governance, and the crisis of the peasantry.
1 A Tale of Two Decades
Recall the finding in the introductory chapter of this book that the fixed-asset investment (FAI) share by the private sector contracted beginning in 1993. Between 1981 and 1989, the private sector's share of fixed-asset investments was 21.4 percent; it declined to 19.8 percent during the 1990-1992 period and then to 13.3 percent between 1993 and 2001. The annual growth of rural private fixed-asset investments slowed to a crawling pace in the 1990s from its torrential pace in the 1980s.
This monumental reversal in the 1990s, affecting the fortunes of hundreds of millions of rural Chinese, has received scant attention. The prevailing view is that in the 1990s, China not only continued but actually deepened the reforms of the 1980s. This is the famous gradualist perspective on China. A central theme of this chapter is that in the 1990s, China reversed many of its productive policies of the 1980s, with real consequences. I show in this section that the income growth in rural China slowed down in the 1990s. In particular, business income growth - the income derived from owning and operating a business - slowed down drastically. I show that in the 1990s, rural Chinese, who otherwise could have entered into entrepreneurship under more propitious conditions, instead "chose" to crowd into the already highly competitive labor markets. They did so despite the fact that not only did the returns from their labor contributions pale in comparison with their business income, they were decreasing as well. The blocking of entrepreneurial opportunities must have been a factor behind this development.
Like the rest of this book, I rely heavily on government reports and documents as the empirical basis from which to draw my findings. The following
114
Capitalism with Chinese Characteristics
excerpts are from a report prepared by a researcher at a think tank associated with the State Council. The researcher, Zhao Shukai (1999), published his findings in the State Council Investigation and Research Report m 1999:
• "For peasants in certain regions, their burdens began to increase in the 1990s."
• In Li village of Henan province, "starting in 1992, most rural households began to experience something new: Their hard-earned money was taken away by the village cadres."
• In Xuantanggang village, "in the 1980s peasants turned over 170 jin of grain per raw to the state; it is now 430 jin per raw."
• "As evidenced by government documents, [the taking of the peasants' belongings] by force began to occur on a large scale in the early 1990s."
The sentiments expressed by Mr. Zhao and Mr. Li quoted at the beginning of this chapter are the tale of the two decades to be told in the following pages. I show that the qualitative assessment reached by Mr. Zhao and Mr. Li are systematically supported by the findings based on large-scale survey data.
1.1 A Reversal of Fortunes
A central theme of this chapter is that the decades of the 1980s and 1990s are extremely different from each other and that economic performance in the 1990s was inferior to that in the 1980s. Any alleged performance differences between the 1980s and the 1990s bear directly on the validity of any explanations about China. The gradualist view of China is supported by data that show China's performance growing stronger over time. The view that China retreated from the reforms will find support in data that show China's performance to be weakening over time.
Let me first cite a finding that directly contradicts my views. It is important to address these data issues up front because we should be explicit about how the data are organized and presented. We again return to Naughton (2007) because it is the most comprehensive collection of the views and data schooled in the gradualist perspective of China. Using the urban and rural household survey data collected by the NBS, Naughton (2007, Table 9.1, p. 210) calculates real per capita household income for rural and urban residents. He shows that between 1985 and 1991, the average rural net household income grew at 2.8 percent per year, compared with 4.9 percent per year on average during the 1991-2004 period. (I mainly focus on the rural areas here, but Naughton also shows in his calculations the superior
A Great Reversal
115
performance of urban disposable income in the 1990s, compared with the second half of the 1980s: 7.7 versus 4.8 percent.)
It turns out that the findings are highly sensitive to how the calculations are periodized. The average figure for the 1985-1991 period includes three years (1989, 1990, and 1991) in which rural households registered the slowest ever rates of income growth in the reform era. In fact, in 1989, rural households registered a negative income growth. The inclusion of these three years in his calculations will necessarily depress the growth rates for the 1980s.
Treating the three years of the Tiananmen interlude as part of the 1980s is problematic. Zhao Ziyang, who presided over economic stewardship of the country in the 1980s, was purged in May 1989 and Jiang Zemin formally assumed the position of Party general secretary in June 1989. It was Jiang, not Zhao, who presided over China between 1989 and 1991. A more analytically accurate approach would be to consider the Tiananmen interlude as a part of the 1990s. Those three years represented a massive retreat, or even an explicit repudiation, of the policies of the 1980s. Credit financing for the private sector was reduced drastically; new restrictions on private businesses -the vast majority located in the rural areas - were enacted; and, as we saw in Chapter 1, rural private fixed-asset investments declined. At a minimum, the three years of the Tiananmen interlude should be considered separately from the 1980s.
If we start with the basic premise that economic policies have a substantial bearing on economic performance, we should categorize Chinese economic performance on the basis of political leadership and economic policies. I do so here on the basis of the same source of data used by Naughton. Panel (1) of Figure 3.1 graphs the annual average growth rates of urban and rural household income against four distinct policy periods. The first policy period is that of rural entrepreneurship between 1979 and 1988 (and the subperiod between 1984 and 1988); the second period is the Tiananmen interlude between 1989 and 1991; the third period is the era of Jiang Zemin and Zhu Rongji; and the fourth period is that of Hu Jintao and Wen Jiabao. The income figures are deflated to their 1978 prices using the official urban and rural income deflators. (Naughton's data are deflated to their 2004 prices.)
A striking pattern in the graph is how low the rural household income growth was during the Tiananmen interlude. Average growth in 1989,1990, and 1991 was only 0.7 percent compared with the double-digit growth during the first eight years of the 1980s. This underscores my earlier point -that classifying the Tiananmen interlude as a part of the 1980s substantially
116
Capitalism with Chinese Characteristics
Panel (1)
14.0% 12.0% 10.0% 8.0%
£
6.0%
Q.
4.0% 2.0% 0.0%
r ■
■ Urban disposable income □ Rural net income
Rural               Rural        Tiananmen Jiang-Zhu era: Hu-Wen era:
entrepreneurship entrepreneurship interlude: 1992-2002 2003-2006 (1): 1979-1988    (2): 1984-1988 1989-1991
Policy Periods
Panel (2)
30 25 20 15
£
o 10
Q.
5 0 -5 -10
■ Net wage income
□ Net household
non-farm business income
1984-1988 19891
1992-2001 2002-2005
Year
Figure 3.1. A great reversal. Panel (1): Average real growth rates of urban and rural per capita household income (based on household surveys, %). Panel (2): Real average annual growth rates of wage and non-farm business income (based on rural household surveys, %). Note: The data for the 1984-1988 period are reported separately here because some analysts believe that the price deflators for the first half of the 1980s are not reliable. Source: Based on rural and urban household surveys, various years.
understates the achievement of the 1980s. A second noteworthy feature of the graph is the discrepancy between urban and rural income growth. In the 1980s, rural income grew significantly faster than urban income; in the 1990s, it was the other way around. Again, as in so many areas of the Chinese economy, the Tiananmen interlude marked the change.
A Great Reversal
117
Rural household income in the 1980s grew at an extraordinarily robust rate. The average growth for the 1978-1988 period was 12.2 percent (after inflation is excluded). This was the decade of entrepreneurship referred to in the previous chapter. A legitimate issue is the reliability of the data. Naughton points out that the data for the earlier period are inaccurate because the rural consumer price index understates rural inflation (and thus overstates growth). This is an important insight. It explains an otherwise paradoxical contraction of income growth in the mid-1980s when rural policies became even more liberal. Whereas there may have been some reduction, the more important reason is the correction in the data series rather than a real policy change.
Let me assume the income growth for the entire decade of the 1980s to be at the rate prevailing in the second half of the 1980s - 7.2 percent rather than 12.2 percent. Even by this conservative estimate, the 1980s were still substantially stronger than the 1990s. During the Jiang Zemin-Zhu Rongji era, from 1992 to 2001, the average growth rate was 4.7 percent, a reduction by 35 percent compared with the rate during the second half of the 1980s. If we look at the entire Jiang Zemin period from 1989 to 2001, the average growth rate of rural income was only 3.8 percent, slightly under half of the growth rate in the second half of the 1980s. During the Hu-Wen period since 2002, growth recovered somewhat, to 5.5 percent. It is important to stress that these growth differences were not a result of the migrant labor income. The net rural household income includes labor income earned by household members working in locations outside the residence of the polled households.
The annual difference between the rural entrepreneurship growth rate of 7.2 percent and the state-led growth rate of 3.8 percent is not an abstract matter. It entails real and substantial income and welfare implications for hundreds of millions of Chinese peasants. This is not only because of the difference in the two growth rates but also because of the extraordinarily long tenure of the Jiang-Zhu leadership - 13 years. If, in the 1990s, the income of Chinese rural households had grown at the rate prevailing in the 1980s - that is, 7.2 percent as opposed to the actual 3.8 percent between 1989 and 2001 - compounded over 13 years, the two rates translated into a massive difference in the levels of rural income. Roughly, a Chinese peasant was 52 percent poorer than he would have been under the lower of Zhao's growth rates (i.e., assuming 7.2 percent for the entire decade of the 1980s).
That rural income growth began to recover under the Hu-Wen leadership is an important finding. For one thing, it illustrates the importance of policy. It is widely known that the Hu-Wen leadership began to address the rural
118
Capitalism with Chinese Characteristics
problems in a proactive manner, even copying the format of the 1980s by issuing consecutive No. 1 policy documents dedicated to rural issues. We can debate whether their measures are adequate to the monumental task -many of these measures were designed to reduce the rural tax rates rather than to augment income growth, an issue I return to in the concluding chapter of this book - but there is no doubt that Hu and Wen take rural issues more seriously than did the Jiang-Zhu leadership.
The recovery of rural income growth under the Hu-Wen leadership is also analytically significant. Some may argue that rural income in the 1990s would naturally slow down after a period of rapid growth in the 1980s. After all, rural income would have had to grow from a higher base in the early 1990s than in the early 1980s. Simple logic says that it is harder to grow from a higher base than from a lower base.
It is extremely important to debunk this view because it attributes the slowdown to a natural economic process rather than to the policies of the 1990s. First, let me emphasize that in 1992, the per capita rural net income was only 449 yuan (in 1978 prices); this was about 81 dollars. It is absurd to believe that income at that low level would cap the growth. The reasoning is also squarely contradicted by the urban data: The level of urban income growth accelerated in the 1990s over the level in the 1980s despite it being from a higher base. The fact that rural income began to recover under the Hu-Wen leadership shows the importance of policy in affecting the growth rates of rural income.
1.2 From Entrepreneurs to Laborers
For rural residents, the most effective way to alleviate poverty is to transition out of low value-added agriculture and into higher value-added industrial or service activities. Theoretically, a Chinese peasant has two options to transition out of agriculture. He can start his own business, either providing services such as buying or selling of agricultural produce or production inputs, or going into the manufacturing of industrial products. Another option is paid employment: A rural resident can become a wage earner by working in a factory or a business owned by someone else. Chapter 2 describes the robust development of both sources of rural industrialization in the 1980s when rural households entered into entrepreneurship on a massive scale and TVEs grew rapidly to absorb the rural labor force.
We know from the previous section that the overall rural household income slowed down considerably in the 1990s. To understand the dynamics of this declining growth, we should decompose the components of the rural
A Great Reversal
119
household income. By 1988, a typical rural household had 545 yuan in net per capita income. Of this, 118 yuan came from wage earnings and 58 yuan came from non-farm business earnings (i.e., profits from operating non-farm businesses). These two non-farm sources of income constituted 21.6 and 10.6 percent, respectively, of the overall household income. Because in the long run, agricultural income was not expected to grow substantially, the growth of rural household income critically depended on the growth of these two non-farm sources of income. Once again, the decades of the 1980s and 1990s differ substantially on this dimension.
Panel (2) of Figure 3.1 presents the annual real growth rates of wage income and non-farm business income averaged over the four policy periods. (Unless otherwise noted, business income refers to non-farm business income in the subsequent paragraphs.) The most visible pattern in the graph is the huge decline in business income growth after 1989. Between 1984 and 1988, growth averaged 27 percent, but during the Tiananmen interlude, growth tanked into negative territory (—0.58 percent). In the 1990s, business income grew in real terms but at a fraction of the rate prevailing in the 1980s (10.5 versus 27 percent). Then, during the Hu-Wen period, growth collapsed completely, to only 0.3 percent. This latest development is very worrisome and evidence that the ameliorative policy measures to solve the rural problems during this time fall far short of addressing the root cause of the problems in rural China - the difficulties faced by rural entrepreneurs to move out of agriculture.
During the 1990s, the business income share of rural income was higher than the level prevailing in the 1980s: 12.4 percent during the 1992-2001 period versus 8.5 percent during the 1984-1988 period. The point here is that the speed at which rural Chinese transitioned out of agriculture through the entrepreneurship channel slowed down in the 1990s, not that non-farm entrepreneurship in the 1990s was at a lower absolute level than that in the 1980s. That said, it is important to point out the following huge difference between the 1980s and the 1990s. In 1984, the non-farm business income share of rural income was only 5.6 percent. Within four years, this ratio almost doubled, to 10.6 percent in 1988. Therefore, the low ratio of the 1980s was primarily a function of the fact that the rural households had started out as completely agricultural producers in the early 1980s and thus they had much distance to go to make the transition to nonagricultural producers. That rural households in the 1990s had a higher business income share than in the 1980s owes largely to the fact that this ratio was already quite high in the late 1980s. Indeed, in the 1990s, this ratio fluctuated widely rather than increasing linearly. The ratio was 13.2 percent in 1993,
120
Capitalism with Chinese Characteristics
10.7 percent in 1995, 14.9 percent in 2000, and then it fell back to 11.5 percent in 2005.
In interpreting these data, let us keep in mind three sets of facts. First, overall rural income growth slowed down substantially in the 1990s compared with the 1980s. Therefore, the share of non-farm business income was rising in the 1990s in part because of the slowdown of overall household income. Second, recall that in 1988, this ratio was already 10.6 percent but in 2005 this ratio was only 11.5 percent, a very small increase. Third, the decade of the 1990s is associated with rapid industrialization and urbanization. Relative to the pace of structural transformation of the Chinese economy in the 1990s, the income composition of rural households experienced only a modest change.
In the 1990s, rural households largely maintained the same growth rate in wage income as they did in the 1980s. This is shown in Panel (2) of Figure 3.1. During the Tiananmen interlude, wage growth was very low. But, during the 1990s, real wage growth recovered to a level close to - but still lower than-wage growth during the 1980s: 9.5 percent during the 1992-2001 period versus 10.7 percent during the 1984-1988 period. During the 2002-2005 period, the growth slowed to 8.33 percent. Still, compared with the huge contrast in the growth rates of business income, these differences between the two decades appear to be small.
That the wage growth kept up while the business income did not provides a keen perspective about the 1990s. In the 1980s, rural residents had two options to transition out of agriculture - they could start non-farm businesses themselves by becoming entrepreneurs or they could contribute labor by becoming workers or paid employees in businesses owned by others. Consistent with our portrayal of the 1980s as an entrepreneurial decade, business income grew much faster than wage income. In the 1990s, the trends were reversed somewhat as the business income growth slowed down considerably, while the wage income stayed at a level close to that of the 1980s.
This was a major development in the 1990s: Rural entrepreneurship as a method to transition out of agriculture was curtailed, whereas transition through labor contributions remained a viable option. The next question is whether this change from rural entrepreneurship to labor participation was a predictable process of a market economy or whether it was the result of policy change. One can think of a perfectly logical explanation for the change - that the returns from labor contributions exceed those from entrepreneurial business income. Rational individuals would then choose
A Great Reversal
121
to allocate more of their efforts toward paid employment rather than toward entrepreneurship.
The differentials between the per capita levels of wage and business income are not terribly revealing about what was occurring. The reason is that in both the 1980s and the 1990s, the per capita wage income always exceeded the per capita business income. In 1985, for example, the per capita wage income was 72.2 yuan, compared with the per capita business income of 32.2 yuan, a ratio of 2.24. In 1995, the per capita wage income was 353.7 yuan, compared with per capita business income of 169.3, a ratio of 2.1. So, clearly, the higher level of wage income relative to business income was a constant in both decades; thus, it cannot explain the change in growth rates between the two decades.
To understand the change in the 1990s, we need a measure that better captures the concept of returns - income earned per unit of labor or capital expended to generate those returns. We turn to a dataset compiled by the Central Committee Policy Research Office and the Ministry of Agriculture (2000). The volume reports on data aggregated from a large-scale survey effort conducted on some 20,000 rural households. The survey, known as the "300 Fixed Rural Observation Villages," commenced in 1984 and tracked the same 300 or so villages every year thereafter. The survey was not conducted in 1992 and 1994 and the published volume provides data only since 1986. We refer to this survey as the fixed household survey hereafter. (The Appendix to this chapter contains more detail about the survey.)
The survey asked the respondents to provide estimates of the number of labor days devoted to the following types of production activities: (1) operating household businesses (broken down by farm and non-farm activities), (2) paid employment activities in the region, and (3) paid migrant employment activities elsewhere. The survey also records streams of income per household attributed to these three production activities. This enables us to calculate the income earned per labor day devoted to a particular activity. To compare data across different years, we deflate all the earnings per labor day to their 1978 prices using the implicit rural income deflators.2
Table 3.1 presents earnings per labor day and the number of labor days devoted to two kinds of non-farm activities - household business and paid employment. Household business is essentially an entrepreneurial activity. An entrepreneur owns and operates the business and his income is the total of all the profits after the business expenses and taxes are deducted.
122
Capitalism with Chinese Characteristics
Table 3.1. Earnings per day and labor-days devoted to household business and paid
employment activities
1986-1988 1989-1991 1992-1999
Panel (1): Household Business
Non-farm business:
-Income per day (yuan)      9.4 yuan               9.6 yuan 12.0 yuan
-Labor-days (% share of   82.6 days (34.2%)    83.9 days (36.1%) 97.3 days (29.0%)
business and paid
employment)
Of non-farm business: Industry
-Income per day (yuan)    12.5 yuan 12.1 yuan                17.0 yuan
-Labor-days (% share of   22.8 days (27.6%) 22.5 days (26.8%)     19.6 days (20.1%) non-farm business)
Panel (2): Paid Employment
Local employment:
-Income per day (yuan)      9.0 yuan 10.3 yuan                 6.0 yuan
-Labor-days (% share of   86.9 days (36.0%) 71.0 days (30.6%)    143.4 days (42.7%)
business and paid
employment)
Migrant employment:
-Income per day (yuan)      3.7 yuan 4.2 yuan 6.8 yuan
-Labor-days (% share of   49.1 days (20.4%)    55.0 days (23.7%)     75.3 days (22.4%)
business and paid
employment)
Note: The fixed household surveys were not carried out in 1992 and 1994. I calculated the values for these two years by taking the average of the two neighboring years. The rural income implicit deflators were used to convert all the values to 1978 prices.
Source: All the calculations are based on data compiled by the Central Committee Policy Research Office and Ministry of Agriculture (2000).
Examples of non-farm businesses include trading, transportation, and industrial production. Paid employment, as the term implies, is employment at a business owned by someone else. The income from paid employment is the wage. We have data on two kinds of paid employment - employment activities within the region of the household and migrant employment activities elsewhere.
Table 3.1 shows that the returns - defined as earnings per labor day -increased for non-farm businesses as a whole, especially for industrial businesses. During the 1986-1988 period, the per-day earnings from non-farm business were 9.4 yuan per rural household; this increased to 12 yuan during the 1990s. (Unless otherwise noted, all monetary figures refer to the
A Great Reversal
123
household average amount per labor day.) This represents a 28 percent increase in real terms. For those households engaged in industrial business, the returns increased even more, from 12.5 yuan per day in the 1980s to 17 yuan per day in the 1990s, an increase of 36 percent. So, clearly, whatever the alleged constraints on rural entrepreneurship were in the 1990s, they did not reduce the returns from entrepreneurship as compared with the level in the 1980s. The returns from labor contributions to local employment establishments declined substantially in real terms. In the 1980s, per-day earnings were 9 yuan; in the 1990s, they averaged 6 yuan, a reduction of 33 percent. Migrant employment earnings experienced healthy growth, from 3.7 yuan to 6.8 yuan.
Putting these findings together suggests the following hypothesis about the 1990s. In the 1990s, there were policies in place that restricted the expansion of rural entrepreneurial businesses. I detail what these policies were in the next section, but credit constraints were among those that increased sharply in the 1990s. The existence of these constraints shows up in the pattern of labor days allocated to business and paid employment activities. Let us look at industrial business earnings. Earnings increased sharply in the 1990s but the time allocations by rural households went in the opposite direction. The number of days allocated to industrial businesses not only declined relatively but also absolutely. In the 1980s, the number of labor days dedicated to industrial business was 22.8 days but, in the 1990s, it was only 19.6 days. The share of labor days also declined, from 27.6 to 20.1 percent.
This is puzzling for several reasons. One is that although industrial business fetched the highest returns among the non-farm business activities, Chinese entrepreneurs in the 1990s allocated an increasing amount of time to other less lucrative activities, such as trading and transportation. The same puzzle exists regarding the time allocation between non-farm business and paid employment. Non-farm business fetched higher returns, and yet it was the paid employment that claimed the highest share of labor time. On a per-day basis, a Chinese rural household in the 1990s earned only half of what it was earning in the 1980s from local paid employment, and yet the same household allocated 143.4 days, an increase from 86.9 days in the 1980s, to this sharply less remunerative activity. The time allocation decisions are systematically mismatched with the relative returns from the various economic activities.
Both of these puzzles can be fairly easily explained by the existence of barriers to rural entrepreneurship. The barriers to entry or to expansion of rural entrepreneurship increased profits to incumbent businesses, thus
124
Capitalism with Chinese Characteristics
explaining why non-farm business income increased in the 1990s. A plausible hypothesis explaining the paradoxical decline of labor days allocated to the higher-margin industrial business compared with the lower-margin service business is that industrial business is more capital-intensive; thus, it is more sensitive to credit constraints than is service business. This is consistent with our aggregate knowledge about China's private sector - that it is highly concentrated in the capital-extensive service sector.
If this reasoning is correct, contrary to the conventional view, rural China in the 1990s suffered from a lack of competition, not from excessive competition. Instead, all the competition converged on the labor markets. Because private businesses and TVEs were constrained from expanding, they created fewer job opportunities and, given that the agricultural income was declining, rural Chinese competed fiercely with one another in order to secure a job. This description is consistent with the wage compression documented previously as well as with the anecdotal accounts of rural Chinese bracing the cruel labor practices, huge wage arrears, and even labor indentures as they navigated the highly unfamiliar urban labor markets.
The other consequence was an increase in the flow of migrants. Because the local employment markets were overcrowded, rural residents sought jobs elsewhere. This latter explanation is consistent with what we know about the 1990s - that rural job migration increased substantially during the decade. This explanation is also consistent with a more recent story that shows that when the income prospects improved somewhat in the rural areas, places such as Guangdong suddenly began to experience a shortage of migrant labor. As Table 3.1 shows, the per-day earnings from migrant employment were not high in the first place, so a modest increase in earnings from other sources, such as agricultural earnings, would tilt the incentives away from migrant labor.
What does all of this mean? First, the rural industrialization that propelled the Chinese economy in the 1980s seems to have taken a very different turn in the 1990s. In the 1980s, the economy was entrepreneurially driven, with the labor deployment split almost evenly between entrepreneurial self-employment activities and paid employment activities. In the 1990s, the rural industrialization continued but it became more of an employment industrialization. The rural Chinese who were exiting from agriculture increasingly crowded into factories or establishments run by others rather than starting their own businesses. They made this "constrained choice" despite the demonstrably lower and decreasing returns from paid employment.
This finding entails enormous welfare implications. An activity that tied up the largest share of nonagricultural labor deployment by Chinese rural
A Great Reversal
125
households actually yielded declining returns over time in the 1990s. This is puzzling, especially given the fact that the labor days devoted to household businesses not only experienced a relative decline compared with other lower-yielding activities, they also experienced an absolute decline over time. A natural question emerges, "Why did Chinese rural households not go into the higher-yielding business operations and instead crowded into the lower-yielding local paid employment?" The answer is that rural China was beginning to be afflicted with increasing financing constraints, and business creation, as an economic activity, was increasingly out of reach for many rural Chinese.
2 What Happened to the TVEs?
We disagree with the inappropriate exaggeration by Comrade Zhao Ziyang on the role of TVEs and with his policy of introducing some unhealthy TVE practices into the large and medium SOEs. But we do not deny the importance of the TVEs. The main challenge facing the TVEs is how to utilize raw materials and inputs in their local rural areas. The markets for the TVEs, except for a few TVEs aimed at the cities, should be primarily in the rural areas, providing for agricultural production and the daily commodities needed by the peasants.
- Premier Li Peng, October 11, 1989 (Li Peng 1989)
It is now widely acknowledged that the TVEs faltered badly in the 1990s. The pace of job creation by TVEs dropped off sharply after 1996. In the 1980s, TVEs had been a primary source of employment and economic growth but, after 1996, the TVEs began to decline steadily. The value added of the TVEs in the share of GDP peaked in 1999 and then leveled off. The financial performance of the TVEs worsened, saddling the rural financial institutions with bad debt. According to survey data, township enterprises as a whole were incurring losses beginning in the mid-1990s (Park and Shen 2000) and today, the very term TVE has almost completely disappeared from the Chinese economic lexicon. The TVEs are history.
Understanding the downfall of the TVEs is critical. China is still a deeply rural society. In 2005, of 778.8 million people classified as active labor force in the official data, 484.9 million were rural employees (or 62 percent of the total). The TVEs were the most dynamic economic force in the 1980s and provided a ready channel for the rural Chinese to transition out of the low value-added agriculture. Their failings in the 1990s had a huge effect on the economic well-being of the rural population. The fate of the TVEs is also a barometer of the policy environment. As I showed in the last chapter, the vast majority of the TVEs were purely private; thus, policies toward TVEs were equivalent to policies toward the private sector. In this section,
126
Capitalism with Chinese Characteristics
I show that the policies toward TVEs became adverse in the 1990s and that it was this change in the policy environment more than those other factors emphasized by economists that explains the demise of TVEs.
2.1 The Downfall of the TVEs
The period from 1978 to 1996 was the "golden era" for TVEs. The period since then has been one of retrenchment. China scholars have provided many explanations for the changing dynamics of the TVEs. There are three prominent strands. One identifies the increasing competition in the market place as an important factor. In the 1980s, according to this explanation, the TVEs were protected from competition. The SOEs had not been reformed and competition from foreign firms was absent. Thus, the rise of the TVEs owed to the specific circumstances of China at the time. Another explanation identifies the public ownership of the TVEs as a problem. The TVEs might have been more efficient than the SOEs, but they still lagged behind the private firms. As private firms began to operate more freely, the position of the TVEs eroded. The third explanation focuses on the change in political incentives. For whatever reasons, in the 1980s, local governments are said to have favored collective firms but, in the 1990s, they began to withdraw their support for collective firms in favor of private firms.
The unstated takeaway of all three of these explanations is that the demise of the TVEs should be appropriately viewed as a sign of progress and deepening reforms. China's maturing marketplace, its more competitive landscape, and the success of its purely private firms are all positive developments. If the demise of the TVEs is the consequence, this is a sign that China is moving forward. Within official Chinese circles, there is another version of this type of reasoning. The TVEs were technologically backward and lacking in brand recognition and international expertise. They were ill-suited for a more modern and technologically advanced China.
In this section, I dispute all these claims about why the TVEs faltered in the 1990s. My own view - and I supply data to support this view - is that the TVEs failed because the national policy environment became inhospitable toward rural entrepreneurship. Before I make this argument, let me first illustrate why some of the commonly accepted explanations for the failure of the TVEs are problematic.
The view that TVEs failed because they were uncompetitive vis-á-vis private firms lacks even surface plausibility. As I detailed in the previous chapter, as early as 1985, an overwhelming number of TVEs were private. By the mid-1990s, it was even more the case that the TVE phenomenon was
A Great Reversal
127
private. In the 1980s, although the private TVEs outnumbered the collective TVEs, their output and employment share was smaller. By 1996, the private TVEs were larger than the collective TVEs by employment, accounting for 56 percent of TVE employment. In terms of output value, in 1987, private TVEs accounted for 23.3 percent of gross industrial output value. By 1997, the share was 51.2 percent. Therefore, in the 1990s, the TVE sector was already substantially private, so it is a rather strange explanation that the TVEs failed because they were not private. The same rationale applies to the explanation that the local support for private firms was the reason for the demise of TVEs. In 1997, 18.8 million TVEs, out of a total of 20.1 million TVEs, were private. There was no better way to support private firms than to support TVEs.
The view that TVEs failed because they were out of step with the rest of the Chinese economy is at best incomplete and at worst misleading. This view somehow casts 18.8 million private TVEs as rigidly wedded to the old business models, incapable of adapting to a new market environment, and unwilling to upgrade their products. Unless one accepts the view that competitiveness is completely exogenous (Is it the water that the rural entrepreneurs drank?), it is reasonable to argue that many of the private TVEs could have gained competitiveness if the business environment had been more hospitable.
In fact, many of them did become competitive, but they tended to be concentrated in one geographic region - the province of Zhejiang (and within Zhejiang, the city of Wenzhou). The private firms based in that province are now designing brands, moving up on the technological chain, and actively investing in R&D. Almost all of these firms started out in the 1980s as private firms registered as TVEs.
The success of Zhejiang points to one explanation for why the TVEs failed in the 1990s that is insufficiently emphasized by Western economists - the increase in credit constraints on TVEs.3 Zhejiang, especially the Wenzhou region, has China's most liberal financial policies toward private firms. In 1999, the short-term bank debt outstanding to the private sector from all financial institutions (including RCCs) was 57.9 billion yuan (People's Bank of China 2000). Zhejiang alone accounted for 11.4 billion yuan of this amount. The other two top provinces were Guangdong (8.4 billion yuan) and Fujian (3.4 billion yuan). Improving competitiveness and upgrading product quality require investments, and investments require capital. This is why Zhejiang firms came to dominate China's corporate landscape and this is why other regions, which repressed their private TVEs through financial and other constraint policies, lagged.
128
Capitalism with Chinese Characteristics
The TVEs succeeded in the 1980s and failed in the 1990s for exactly the same reason - that they were substantially private. It was the national policy environment that changed between the two decades. In the 1990s, mainly in the rural areas, it became increasingly difficult to obtain financing. This was in part because of a retreat from the financial liberalization of the 1980s and also in part because the TVE policies changed. The quote from Li Peng at the beginning of this section provides a useful way to reconstruct the TVE policies of the 1990s.
Li Peng's prescription for the TVEs is that they should primarily aim at the rural markets. (To the extent that the view of Chinese elites on TVEs - that they were low-tech - held any truth, it was a result of deliberate government policies tying the TVEs to low-tech rural China.) This perspective on TVEs constituted a dramatic shift from the 1980s. In the 1980s, the reformers viewed the TVEs as a source of competition to the urban-based SOEs and as a platform for rural residents to transition themselves out of low value-added agriculture. (Presumably, these are the unhealthy practices referred to by Li Peng.) Completely contrary to the view held by many Western economists, the reformers in the 1980s actively encouraged the TVEs to compete with the urban firms, rather than insulating them in an exclusively rural environment.
In fact, the reformers went even one step further. They actively encouraged the urban firms to establish direct production and sourcing links with the TVEs. In 1978, the government directed urban factories to shift their procurement to rural enterprises and to transfer technology to them. By the mid-1980s, accordingto Naughton (1996, p. 155), many regions in Jiangsu-still rural in the 1980s - had a significant amount of subcontracting work from Shanghai. One of the earliest beneficiaries of this policy was Wanxiang Auto, now China's largest private automobile-component producer. In 1980, Wanxiang was selected by the China Automotive General Corporation as one of three suppliers of gears. This was a turning point in the development of the firm (Wu Xiaobo 2006, p. 57).
Li Peng's criticism of the TVEs in the 1980s was not idle talk. Nor was it a temporary policy departure in the aftermath of the Tiananmen crackdown. In fact, the 1997 TVE Law, promulgated eight years after Li Peng made those remarks, closely reflects Li Peng's thinking. In his 1989 speech, Li Peng viewed the TVEs as agricultural businesses, mainly serving rural areas and markets. Article 2 of the 1997 TVE Law defines the TVEs as those enterprises "undertaking to support agriculture" (Editorial Committee of TVE Yearbook 1997). The second marker laid down by Li Peng was the emphasis on the collective ownership of TVEs. Again, the 1997 TVE Law
A Great Reversal
129
reflects this thinking. Article 4 spells it out clearly: "TVE development must observe the principle of rural collective ownership as the main form, together with economic diverse ownership forms." The 1997 TVE Law also sanctioned county governments to establish "TVE development funds" to invest in the TVEs.
In the 1990s, the liquidity constraints on the TVEs were substantial. In 1993, the state banking system allocated 16.8 billion yuan for all TVEs. But the Ministry of Finance believed that the TVEs needed 200 billion yuan each year (Ministry of Agriculture 1995). There was also another source of liquidity constraints. China scholars have documented that the SOEs operated with increasing losses since the late 1990s. One effect was that the SOEs accumulated a large sum of accounts payable. It has been estimated that the accounts payable to the TVEs was 381.3 billion yuan in 1994 and 494.2 billion yuan in 1995. To put these numbers in perspective, the entire bank credit to the TVEs in 1994 was just 368.6 billion yuan (Ministry of Agriculture 1996, p. 14 and p. 2-2). The effects of the credit curbs were instantaneous. According to an internal report prepared by the Ministry of Agriculture, in 1994 the number of private TVEs fell by 21.5 percent and their employment fell by 10.8 percent. All of the increases in the TVEs occurred at the collective end. Township and village TVEs increased both in number of establishments and in employment (Ministry of Agriculture 1995).
2.2 TVE Privatization
Many Western economists believe that the private sector grew vigorously in the 1990s because the government increasingly came to support the privatization of TVEs. In the 1990s, the purely ideological opposition to private ownership eased considerably, but in rural China private-sector development was not as straightforward as many have assumed. For one thing, it is important to keep in mind the fact that the vast majority of the TVEs had started out as private businesses in the first place. In terms of establishments, as I showed in the last chapter, as early as 1985 more than 10 million out of 12 million TVEs were private. One significant development in the 1990s was that the coastal and richer provinces began to privatize their TVEs. Because these provinces have a large weight in the national data, the private share of TVEs grew in terms of output and employment.
It is important to address this view that much of the TVE privatization occurred in the 1990s because it fits with the gradualist interpretation that Chinese reforms got deeper over time. My argument is that this view
130
Capitalism with Chinese Characteristics
overstates the importance of the decade of the 1990s in terms of TVE privatization. The overstatement comes in two forms. One is that this view uses the rising number of one type of TVE, known as a shareholding cooperative (gufen hezuo qiye), in the 1990s as evidence of increasingly liberal privatization stance. In reality, many of the shareholding cooperatives grew organically from smaller private rural businesses rather than as a product of privatization of collective TVEs. Second, many local governments did privatize in the 1990s, but it is not true that local governments did not privatize TVEs in the 1980s.
Shareholding cooperatives are essentially employee-owned firms and, by the late 1990s, many of the TVEs took this particular corporate form. Some of the shareholding cooperatives were no doubt converted from the collective TVEs, but many others were, in fact, an organic outgrowth of household businesses. In Wenzhou, by 1987, the shareholding cooperatives were already quite sizable, accounting for 47.9 percent of the total TVE output value. Of the 15,000 shareholding cooperatives, two thirds actually evolved from household businesses and one third were a result of TVE privatization (Editorial Committee of TVE Yearbook 1989b, p. 346). A 1987 Politburo document provides two rationales for the shareholding cooperative experimentation. One is that it led to a separation of government and enterprise management and the other is that it helped private enterprises raise capital (Editorial Committee of TVE Yearbook 1989b, p. 520). The second rationale has nothing to do with privatization; it is about helping household businesses that have attained scale to raise capital.
Another issue about the shareholding cooperatives is that the number of shareholding cooperative TVEs did not increase linearly. The peak was reached during the 1993-1994 period after Deng's Southern Tour and before many of the financing constraints on the rural private sector were institutionalized. This was the most liberal period in the 1990s. But, after the initial growth between 1993 and 1994, the number of shareholding cooperatives declined sharply in 1995 and 1996. According to the Ministry of Agriculture (1995), in 1994, there were 200,000 shareholding cooperatives, a rapid increase from 130,000 in 1993. But, in 1995, their number had decreased to 182,427 and then to 143,477 in 1996. The number of township/village-level TVEs also declined during this period but less rapidly. Therefore, the ratio of shareholding cooperatives to collective TVEs started at 7.7 percent in 1993, rose to 12.2 percent in 1994, and then fell to 9.3 percent in 1996.
Western academics believe that TVE privatization occurred largely in the 1990s (Whiting 2001, pp. 289-290). In their survey conducted in Jiangsu and Zhejiang, Brandt, Li, and Roberts (2005) report that cumulatively between
A Great Reversal
131
1993 and 1998, 34.2 percent of the TVEs were privatized and another 15.4 percent were shut down. The authors link the TVE privatization to "a fundamental shift in central government policy" in the early 1990s (Brandt, Li, and Roberts 2005, p. 525). Similarly, Oi (1999) reports on the privatization of TVEs in Shandong province in the 1990s.
In fact, what these academics have documented might very well be the tail end of the TVE privatization, not the beginning. Jiangsu is a well-known bastion of collective TVEs and it was among the last to privatize, not a pioneer. Zhejiang also privatized many of its collective TVEs in the 1990s. However, the first wave of the TVE privatization occurred in the 1980s and mainly in the poor, interior provinces, and it was poorly documented. As I showed in the last chapter, the share of private TVEs in the total TVE output was much higher in the poorer and more agricultural provinces than in the richer and more industrialized provinces. Coastal provinces might have finished the TVE privatization; they did not start it.
This empirical detail is important because it helps us assess policy developments in the 1980s and in the 1990s accurately. The explicit policy endorsement of TVE privatization happened much earlier than asserted by Brandt, Li, and Roberts. The first national policy endorsement that I found is a major Politburo document entitled, "Deepening rural reforms," issued on January 22, 1987. Article 4 of the document states, "small [rural collective] enterprises can be leased or sold to individuals to operate" (Editorial Committee of TVE Yearbook 1989b, p. 519). But, this 1987 document acknowledged and sanctioned the apparent large-scale privatization of collective TVEs already underway. Although the evidence is scattered, the poor performance of the collective TVEs at the very start of the rural reforms had already prompted a spontaneous wave of privatization of these loss-making firms.
As early as 1987, in Wuxi of Jiangsu province - the progenitor of the Sunan model - 100 collective TVEs had been acquired by other firms; 200 were leased out to managers, and 11 were converted into shareholding companies (Editorial Committee of TVE Yearbook 1989b, p. 328). If a stronghold of collective TVEs already began to privatize as early as 1987, it is only logical to assume that privatization might have occurred on a larger scale elsewhere. In fact, the rapid growth of private TVEs documented previously was, in part, a product of the privatization process. It was their initial source of financing.
According to the World Bank researchers, as many as 50 percent of the private entrepreneurs in the poor regions of the country got their startup capital from seizing control of collective fixed assets. According to Lin (1990, p. 177), village governments "sold some or all of their enterprises to
132
Capitalism with Chinese Characteristics
individuals at extremely low prices, leased them at a fixed rent, or contracted them out in return for a certain percentage of after-tax profits." In Jieshou, "wholesale" privatization and "outright" sale of collective TVEs occurred as early as 1979-1980. In one township, every single collective TVE was privatized and in another township, half of the collective TVEs were privatized. In Shangrao county, the share of private TVEs rose from 0 to 50 percent, in just three years from 1983 to 1986, through both the organic growth of the private TVEs as well as privatization of the collective TVEs (Byrd 1990).
Chinese sources also report on numerous instances of collective TVE privatization during the 1980s. Editorial Committee of TVE Yearbook (1989b, p. 222) notes the transformation of "originally commune and brigade enterprises" into shareholding cooperatives through employee stock options and new share issues in Sichuan, Shaanxi, Shanxi, and Henan. An investigation into Qinhe county in Hebei province states, "In that region, there were not many collective enterprises in the first place. Of the ones that were collective, the majority were already divided among households" (Editorial Committee of TVE Yearbook 1989b, p. 232). In Wenzhou city, by 1987, more than 4,000 collective enterprises were converted into shareholding companies through either outright privatization or new share issues (Editorial Committee of TVE Yearbook 1989b, p. 346). In Zhejiang province, in 1988, 605 collective TVEs were sold through auctions, 2,210 were leased out, and 364 were acquired (Editorial Committee of TVE Yearbook 1989a, p. 40). An article on shareholding cooperatives details the various privatization practices in Shandong, Shanxi, Shaanxi, Hebei, Wuhan, and Liaoning. The practices ranged from dilution of government shares to outright sales. In Hebei, these various ownership experiments involved some 14,000 firms (Editorial Committee of TVE Yearbook 1989a, pp. 154-155).
The privatization of the collective TVEs financed the initial startup of the private firms. This observation turns on its head the idea that the development of TVEs owed their legacy to the Great Leap Forward. Lin (1990, p. 177) gets the causal direction right when she remarks, "So, in a sense, failure in the management of the original TVCEs [township and village collective enterprises] was a factor in the successful development of today's private enterprises." Here is an ironic twist to the conventional wisdom of the TVEs: To the extent that the collective legacy of the Great Leap Forward contributed to the TVE development in the 1980s, it was through the failure, not the success, of the collective TVEs.
In the 1980s, private TVEs expanded from both rapid organic growth as well as privatization of collective TVEs. This observation suggests that we
A Great Reversal
133
need to rethink about a common view in the writings of the Chinese economy - that TVEs owed their lineage to the collective TVEs from the Great Leap Forward period. This view errs on the facts because most of the TVEs were founded during the reform era. The view also errs on the reasoning. It is predicated on the belief that collective TVEs were successful.4 When the TVEs are segmented correctly - into their true ownership categories such as township, village, or private categories - the evidence is very clear that the private TVEs were more efficient than the collective TVEs. For example, township TVEs reported heavy losses in the World Bank TVE study (Wang Xiaolu 1990). Nationwide statistics for 1985 show that the average output, profits, and wages of private TVEs were between 50 and 70 percent higher than those of comparable collective TVEs (Lin 1990, p. 181). The collective TVEs appear to have begun to incur losses as soon as the reforms began. Even in Wuxi, a region known for the best-performing collective TVEs, 11 percent of the township TVEs suffered losses. The much-touted Sunan model actually began to flounder at the very start of the reform era. Between 1980 and 1984, the collective TVEs in Jiangsu experienced a plunge in after-tax profits by 25 percent (Zhang Yi 1990, p. 192).5
2.3 The Resurgence of Collective TVEs
An additional source of complications in interpreting the TVE privatization in the 1990s is that in the 1990s, there was also a resurgence of collective TVEs. However, this resurgence of the collective TVEs occurred in the backward, small provinces; thus, this development did not show up in the national data. In the 1980s, there was a standard reference to government policy on TVEs - the "four-wheel drive" policy (silunjuedong). The four wheels here refer to the four levels of TVEs: (1) townships, (2) villages, (3) alliance and multihousehold businesses, and (4) households. The first two represent the collective component of the TVEs and the last two represent the private component. Throughout the 1980s, the standard policy formulation was to support the development of all four "wheels" of the TVEs. For example, the 1987 policy document adopted by the Politburo, "Deepening the rural reforms," uses the phrase "simultaneous" development (yiqi shang) of four wheels (Editorial Committee of TVE Yearbook 1989b, p. 520). I know of no policy document in the 1980s that explicitly differentiated among the four different types of TVEs; quite the opposite. Numerous policy documents went out of their way to stress equal treatment of different types of TVEs.
This was to change in the 1990s. A speech by the Minister of Agriculture on January 5,1990, portended a new policy formulation. In this speech, the
134
Capitalism with Chinese Characteristics
Minister revised the so-called four-wheel policy and added that the collective TVEs were the mainstay of the TVE sector (Editorial Committee of TVE Yearbook 1990, p. 8). One policy document even invoked a term customarily associated with the nationalization campaign of the mid-1950s to describe the policy visions for the TVEs.6 The reformulation of the TVE policy to favor collective TVEs was reiterated throughout the 1990s, not just in the immediate ideological aftermath of the Tiananmen crackdown. Several articles in the 1997 TVE Law stress the primacy of the collective TVEs. Investment allocations closely follow the new policy formulation as well. Chapter 1 shows that fixed-asset investments in the rural collective sector grew rapidly in the 1990s. During the 1993-2001 period, rural collective fixed-asset investments grew at an annual rate of 9.1 percent compared with rural private-sector fixed-asset investments growing 7.5 percent.
Although the overall size of private TVEs increased in the 1990s, there are two important observations. One is the rate of this increase. In 1989, private TVEs already accounted for 50 percent of TVE employment. This ratio remained roughly constant, around 51 percent, until 1994. Only after 1994 did the private share of TVE employment begin to rise significantly above the level prevailing in the late 1980s.
The second observation is that this overall rise in private TVEs at the national level masks a huge variation at the regional level. This was not just a variation in the speed of the rising private TVEs but also a variation in the direction of private TVE development. In some regions, private TVEs actually declined in the 1990s relative to the size of the collective TVEs. In 1987, at the national level, private TVEs accounted for 32.1 percent of the gross output value; 10 years later in 1997, the share was 51 percent, a gain of 19 percent. Against this overall increase, however, seven provinces experienced a decline or stagnation of the share of private TVEs. The greatest decline occurred in Heilongjiang province: In 1987, private TVEs already accounted for 47 percent of TVE output; by 1997, this share had declined to 19.5 percent, a reduction of 27.5 percent. The six other provinces are (with the reduction magnitude in brackets) Guizhou ( — 17.3 percent), Qinghai (-16.6 percent), Hebei (—3.2 percent), Henan (-0.6 percent), Beijing (-1.4 percent), and Anhui (0 percent).
We do not know what happened in these seven provinces but it is important to understand the implications. One is that these seven provinces are relatively small in terms of their total weight in the national data on TVEs. In 1997 Anhui, with the largest weight, accounted for 5 percent of the national total. All the others accounted for less than 3 percent. Because private TVEs
A Great Reversal
135
expanded in those provinces with a large TVE sector, national data show an increase in private TVEs. Zhejiang, which contributed 9.8 percent to national TVE output, increased the private share of TVEs by 35.8 percent. Jiangsu, another large TVE province, increased the private share of TVEs by 17.7 percent. The expansion of the private TVEs in these large TVE provinces masks the retrogressions in the smaller TVE provinces in the national data.
Private TVE development is not a mere statistical matter. It entails real welfare consequences. Provinces such as Jiangsu and Zhejiang, the two coastal, richer regions in China, had favorable endowment factors - such as access to FDI, trade, and an urban economy - to fall back on. It was the poor regions of China that most needed indigenous, bottom-up entrepreneurship because they lacked alternative means of economic development. We saw in the last chapter that rural residents in poor provinces such as Guizhou went into entrepreneurship to improve their standard of living. Except for Beijing, the agricultural population represented about 80 percent of the total in those provinces in which private TVEs contracted. They also had a high concentration of poverty. So, the welfare implications of their lagging private-sector development were substantial.
With more data available in the future, we ought to revisit this period and try to understand exactly why the rural private sector contracted in these poor provinces. Here, let me provide another corroboration of the lagging private-sector development in China's poor provinces. Recall the finding in the World Bank TVE study that in the mid-1980s, the private sector was already substantial in size in two of their poorer research sites, Jieshou county of Anhui province and Shangrao county of Jiangxi province. In the richer research sites - Nanhai of Guangdong and Wuxi of Jiangsu, the private sector was relatively small. The World Bank researchers reasoned that this was an organic result of the respective endowment factors. The richer regions started out with a more developed collective sector and they naturally gravitated toward collective mechanisms of economic and industrial development. The poorer regions never had the luxury of the choice. Private-sector development became a default mechanism for the economies of these regions to grow.
This highly convincing explanation, however, makes the policy making look easier than it was in reality. A vital condition is necessary to enable the process of natural selection to work - policy makers have to accommodate themselves to the economic reality on the ground rather than to forcibly impose their own visions. My contention is that the latter occurred on a large
136
Capitalism with Chinese Characteristics
Table 3.2. Percentage shares of gross industrial output value by private TVEs in four
regions: 1985 and 1998 (%)
			Based on NBS Industrial	
	Based on World Bank Survey:		Firm Dataset: Large Firms Only	
	Firms of all Sizes		(Sales >5 million yuan)	
	1985	1986	1998	
	(la)	(lb)	(2a)	(2b)
	Private share	Private share	Private share	Private share
Regions/TVE indicators	of TVEs	of TVEs	of TVEs	of all firms
Jieshou (Anhui)	51		12.4	3.95
Shangrao (Jiangxi)	35		24.6	9.5
Nanhai (Guangdong)	10	12	17.1	15.1
Wuxi (Jiangsu)	3		13.2	9.7
Note: The private firm classification is based on the Guangdong definition; i.e., registered private-sector firms plus those nonstate firms with individual share capital exceeding 50 percent. Sources: The 1980s data draw from Table 9.1 in Byrd (1990, p. 195). The 1998 data are based on the NBS industrial firm dataset.
scale in the 1990s - that the local governments in the poor regions, instead of facilitating a natural, organic process of private-sector development, poured financial and other resources into the collective sector. Table 3.2 presents the estimates given by the World Bank TVE study of the share of private TVEs in the four counties in the mid-1980s. Consistent with the view put forward in this book that the poorer regions in the 1980s pioneered private-sector development, Jieshou and Shangrao counties - the two poorer research sites in the World Bank TVE study - had a higher share of private TVEs than the two richer research sites, Nanhai and Wuxi. In the 1990s, the situation reversed itself. Our data for the 1990s come from the NBS industry census and the year is 1998. The census data, which cover larger firms above 5 million yuan in sales, show that the private shares of TVEs declined in Jieshou and Shangrao but they increased in Nanhai and Wuxi.
Given that the collective TVEs began to incur losses as soon as the rural reforms began, it is implausible that the collective TVEs gained against the private TVEs in Jieshou and Shangrao as well as in the aforementioned seven provinces in the 1990s due to their efficiency and dynamism. Instead, they gained market shares in the 1990s due to policy support. These policy developments were hugely destructive. The collective TVEs wasted the resources allocated to them. They might have gained market shares against the private TVEs located in the same regions but because the resources went
A Great Reversal
137
to the less competitive TVEs, compared with the country as a whole, the TVEs in those provinces lost market shares. Guizhou, Henan, Hebei, Hei-longjiang, and Guangxi, the poor provinces that experienced a contraction of the share of private TVEs between 1987 and 1997, also experienced a contraction in their share of TVE output - inclusive of both collective and private TVEs - in the national total. The TVEs of Henan accounted for 6.62 percent of the national output of TVEs in 1987; in 1997, the share was 3.54 percent. Guizhou's share declined from 0.55 percent in 1987 to 0.38 percent in 1997. In the case of Hebei, it went down from 6.6 percent in 1987 to 4.55 in 1997.
We still have one piece of the puzzle to solve before we can close the loop on the subject of the TVEs. If the private TVEs lagged in some of the poorest provinces, why did they grow in the richer provinces? Jiangsu province, the progenitor of the collective TVE model, privatized many of its TVEs in the 1990s. By 2004, even among the largest TVEs, individual share capital was very important, accounting for 47.5 percent of the total share capital (Ministry of Agriculture 2005). A plausible hypothesis centers on the role of industrial policy. During the 1990s, the Chinese state adopted a policy platform officially known as "grasping the big and letting go of the small." "Grasping the big" means policy support for the large incumbent firms and "letting go of the small" means privatization of small firms.
Here is how this policy approach might have led to the divergent developments between the rich and poor provinces in terms of TVE development. The most valuable and the largest assets in the rich provinces resided in the traditional state sector, rather than in the TVEs. Thus, the logical approach in those regions was to restructure the SOEs, often by massive fresh investments and/or by forming alliances with FDI. The TVEs in these regions were small relative to the incumbent SOEs and were thus relegated to the privatization part of the policy program (i.e., "letting go of the small").
The poor regions had entirely different endowment conditions. They had a relatively under-developed state sector (and this is the reason why the private sector was allowed to develop there in the first place). They also had a paucity of FDI supply, which precluded this particular restructuring option. Their incumbent large firms comprised collective TVEs, which were then targeted for support under the policy of "grasping the big and letting go of the small." So, ironically, exactly the same dynamics behind the rise of private TVEs in the poor provinces in the 1980s then explains the endurance and even the resurgence of collective TVEs in the 1990s - the absence of viable developmental alternatives.
138
Capitalism with Chinese Characteristics
3 The Great Financing Squeeze
Under the banking regulations, individuals are not allowed to engage in financial operations. The emergence of private (siren) credit shows that our financial work falls short of what is needed. This requires that our credit cooperatives and agricultural banks improve their services. This is a huge task.
- Chen Muhua, governor of the People's Bank of China, January 31, 1986 (ChenMuhua 1987, p. 105)
Those funds, mutual assistance associations, savings associations, capital service departments, share capital service departments, fund clearing centers, and investment companies established prior to this order and operating above the state law should be restructured with a deadline according to the regulations of the State Council. Those entities that operate after the deadline and continue to engage in illegal financing should be stamped out according to this order. Those with serious violations of a criminal nature should be held accountable for their legal responsibilities.
- An order from the State Council to ban all illegal finance (State Council 1998)
The first quote comes from Madame Chen Muhua, who was the governor of the People's Bank of China (PBoC), China's central bank, between 1985 and 1988. In this quote, Madame Chen, viewed by many Western journalists as conservative in outlook and wooden in character, was using private - and essentially illegal - financial transactions as a benchmark for the state-owned financial system. In her judgment, China's formal financial system was not up to the task and she urged it to reform. In other speeches given between 1985 and 1987, she constantly implored the state-owned financial institutions to do a better job - whether to draw deposits or to provide loans - in order to compete with private financing. Several times, Madame Chen held up Wenzhou - the bastion of capitalism in China - as a model to be emulated by the rest of the country.
Chen Muhua used the term siren to describe some of the financial practices. Two Chinese terms connote the idea of private ownership. One is siren, meaning private or individual; the other is minjian, literally meaning among the people or nongovernment. Siren is more overtly private and thus more ideologically sensitive than minjian, although the two terms do not differ conceptually. So, there is a preference for using minjian rather than siren in Chinese political discourse. But Madame Chen did not shy away from using siren. In fact, Chinese financial officials went even beyond Madame Chen. A statement by Han Lei, president of the Agricultural Bank of China (ABC), used the term siren as early as 1984 in a discussion on the direction of the financial reforms. Bank documents in the 1980s were peppered with
A Great Reversal
139
references to siren or minjian when discussing bank reforms. In the 1980s, rural China experimented with substantial financial liberalization, the main elements of which were (1) adoption of an accommodating and supportive credit policy toward the private sector by state banks, (2) the proliferation of informal financial instruments, and (3) tacit permission for informal financial instruments exclusively servicing the private sector.
With an ever-increasing intensity, much, if not all, of the financial experimentation in the 1980s was terminated or completely reversed beginning in the second half of the 1990s. The 1998 State Council order, whose stern warning is quoted in part at the beginning of this section, is extremely telling of the winds of change in the 1990s. Rather than viewing the informal finance as a useful complement to the official finance, the Chinese state began to systematically stamp out those providers of capital outside the state banking system. The government began to curb the operations of nationwide semi-official financial institutions, rural cooperative foundations (RCFs), in 1993 and completely banned their operations in 1998. In the 1990s, the two terms siren and minjian completely disappeared from bank documents (except when announcing bans on private financial transactions).
The crackdown on informal finance was both determined and ferocious. In 1991, an illiterate housewife in Wenzhou paid the ultimate price - Zheng Lefang was executed for "financial fraud." Zheng personified the turning point in China's financial policies. She had committed her alleged crimes in 1986 but she was not executed until 1991 (Wu Xiaobo 2006, p. 175). In the 1990s, numerous rural entrepreneurs who had been forced to tap into or organize underground financing because of the massive inadequacies of China's banking system were arrested. A famous case involves Mr. Sun Dawu, a rural entrepreneur who ran an animal feed company in the impoverished province of Hebei.7 In May 2003, Mr. Sun was arrested for "illegally absorbing public funds." Mr. Sun had refused to bribe bank officials to obtain loans. Instead, he turned to the employees of his company and asked them to contribute funds. This practice, widespread in the 1980s and a legitimate source of start-up capital for many TVEs, now ran into the iron fist of the Chinese financial regulators determined to stamp out all forms of informal finance. Sun's company was destroyed. (In his prison cell, Mr. Sun coined a phrase thereafter invoked by many Chinese journalists, "Chinese peasants, your name is misery")
Western scholars are keenly aware of the inadequacies of the Chinese banking system.8 In 1998, Nicholas Lardy argued that the Chinese reforms were unfinished because the financial system was unreformed (Lardy 1998). Other researchers as well have reported on backpedaling of the financial
140
Capitalism with Chinese Characteristics
reforms. For example, Park and Shen (2001) note that authority to issue new loans became highly centralized during the course of the 1990s and a study by the International Finance Corporation, based on a survey in the late 1990s, shows that newer private firms faced greater financing constraints than older firms (Gregory, Tenev, and Wagle 2000).9 Other studies have reported on the deteriorating rural finance in the 1990s (Nyberg and Rozelle 1999; International Fund for Agricultural Development 2002).
I would agree with all of these assessments but I go one step further. My argument is that China reversed many of the productive and innovative financial practices it had adopted in the 1980s. In this section, I first present evidence that rural finance became very constraining in the 1990s. I then contrast the financial policies in the 1980s with those in the 1990s. The primary empirical basis to determine China's financial policies in these two decades is the thousands of pages of bank documents in 22 volumes.
3.1 The Poor State of Rural Finance
Recall the finding in Chapter 1 that rural private fixed investments grew rapidly in the 1980s. Fixed-asset investments are typically heavily financed by external sources of capital - bank loans or new share issues. It is not unusual that the construction of a new production facility is 50 to 70 percent financed by outside capital.
The rapid growth of rural fixed-asset investments in the 1980s illustrates a phenomenon virtually unknown in the West - the supply of bank capital to the private sector in the 1980s was plentiful. The ample supply was a function of two developments. One was a dramatic policy shift by Chinese banks toward a more business-friendly stance and more supportive of private-sector clients. The other was substantial financial liberalization, defined as those policy measures that made control of existing financial institutions more private and allowed private players a greater role in providing financial intermediation services. On both fronts, China moved backward, rather dramatically, in the 1990s.
Survey research undertaken in the 1980s shows a surprisingly high level of loans provided to private entrepreneurs when they first started their businesses. (Data on loan availability during the operating stage are scarce.) Two Chinese sociologists, Zhang Houyi and Ming Lizhi, summarize the findings from six large-scale surveys conducted in 1987 (Zhang Houyi and Ming Lizhi 1999, Table 9, p. 55). One survey, covering 97 firms in 11 provinces, shows that 40.6 percent of the start-up capital came from bank loans. (Unless otherwise noted, bank loans here refer to those funds made
A Great Reversal
141
available to the firms when they first started.) For 281 firms in Hebei, the ratio was 54.8 percent, for 56 firms in Hunan the ratio was 28.5 percent, for 130 firms in Shaanxi it was 66.3 percent, for 10 firms in Guangdong it was 34 percent, and for 50 firms in Wenzhou it was 23.3 percent. The average ratio in these aforementioned six surveys is 41.3 percent.
The World Bank TVE study, referenced in the last chapter, also reports very high levels of credit availability to the private sector. Lin (1990, fn. 3, p. 188) reports on a survey of 56 private firms in Tianjin in 1985. Of those firms with a total investment of less than 50,000 yuan, bank loans accounted for 38.8 percent of their funds; of those with an investment between 50,000 and 100,000 yuan, bank loans accounted for 43.6 percent; and of those firms with investments of more than 100,000 yuan, bank loans accounted for 69.9 percent. One of the World Bank researchers, William Byrd (1990, p. 209), thus observes, "Banking institutions already see well-established private enterprises as solid borrowers." Byrd also reports that local banks that lent heavily to private-sector firms had lower non-performing loan (NPL) ratios.
It is definitely not true that private entrepreneurs in the 1980s were unable to access bank loans. But did access to loans become more or less difficult in the 1990s compared with the 1980s? To compare the two decades directly, we go to three sources of information that organize and report data on a consistent basis for both the 1980s and 1990s. A head-to-head comparison shows that private-sector access to finance, especially in rural China, was substantial in the 1980s, but it became extraordinarily constrained in the 1990s.
The first source of information is the fixed rural household survey we used to demonstrate the changing labor time allocation of Chinese peasants in the 1980s and 1990s. The fixed rural household survey provides data on loans obtained from banks and rural credit cooperatives (RCCs) from 1986 to 1999. To see the trends over time, I deflated the bank loans to their 1978 prices using the rural price index. In 1986, an average rural household obtained 84.2 yuan from banks and RCCs. This rose to 99.5 yuan in 1987 and to 92.3 yuan in 1988. Then, it contracted to 52.3 yuan in 1989. From that point on until 1999, the average rural household bank loans in real terms never exceeded their 1987 level. The peak year of the 1990s was 1996 when the average rural household bank loan was 92 yuan; in all other years, the figure was either below 80 yuan or only slightly above. Only in 1999 did the level exceed the peak reached in 1987. In that year, the amount of bank loans was 103 yuan.
In absolute terms, the average amount of formal loans per household did not increase in the 1990s compared with the 1980s. In relative terms, it declined. Because we are mainly interested in the role of bank loans to help
142
Capitalism with Chinese Characteristics
Chinese peasants transition out of agriculture, we compare the amount of bank loans with the amount of non-farm operating income. In 1986, 84.2 yuan of bank loans represented 28.3 percent of the operating income from non-farm sources. For the next three years, this ratio remained above 28 percent. In the 1990s, the ratio declined on a continuing basis. By 1999, the ratio was only 21.9 percent, a 20 percent reduction from the 1986 level. (Later in this chapter, I present data on the supply side to show that a main source of financing in rural China, the RCCs, shrank to a point of total irrelevance.)
Our second source of information is the private-sector surveys we used in Chapter 2 to ascertain the rural origins of Chinese capitalism. Question 8 in the PSS2002 asked the respondents to select their sources of startup capital from the following sources: (1) savings from running small businesses, (2) savings from running small-scale productions, (3) donations from friends and relatives, (4) wages, (5) informal loans, (6) bank loans, and (7) inheritances. Let me compare the number of firms that checked off bank loans versus those that checked off informal finance in their responses. The PSS2002 contains information on the year in which the firm was founded so we can compare the responses to this question in the two decades. Because very few firms in the PSS2002 were established before 1984,1 exclude those firms in the data analysis. Also, I report on the findings only on rural firms, although the findings on the entire sample do not differ.10
During the 1984-1989 period, 32.6 percent of rural firms reported receiving bank loans in the first year of their business. The highest ratio was in 1985 when 50 percent reported receiving loans. The year 1987 was also high, at 38.5 percent. During the 1990-2001 period, this ratio declined sharply, to 26 percent. Some years show very low numbers. For example, in 2001, the year often touted as an ideological breakthrough for China's private sector when Jiang Zemin unveiled his doctrine of the "three represents," only 13 percent of rural private firms received bank loans. The highest ratio was in 1999 when 34.6 percent of rural firms received bank loans, but this is nowhere near the 50 percent already reached in 1985.
As we already saw in the rural social-economic survey, informal finance skyrocketed to meet the unfulfilled credit demand. We find exactly the same dynamics in the PSS2002. During the 1990-2001 period, 29.7 percent of rural firms reported receiving informal loans, as compared with 26.3 percent during the 1984-1989 period. Thus, formal finance was more important to rural firms in the 1980s than it was in the 1990s; in the 1990s, the importance of these two sources of finance were reversed, with informal finance surpassing formal finance as a source of start-up capital.
A Great Reversal
143
One interpretation of the emerging role of informal finance is that the government became more tolerant of private providers of capital. Thus, it can be a sign of openness, not of closure, as Tsai (2002) explains the prominence of informal finance in some regions of China. This is the right perspective to explain the 1980s when the reformers endeavored to make the state financial institutions cater to private entrepreneurs and to allow a degree of opening and competition in the financial sector. The formal and informal sources of finance complemented one another.
In the 1990s, rather than being complementary, the formal and informal sources of finance became substitutes for one another. The authorities oriented the banking system away from the private sector; thus, the credit constraints on the private entrepreneurs drove them to rely more heavily on informal finance. The way to distinguish the substitution and complementary relationship between the two sources of finance is to look at how they relate to one another. In the 1980s, the reliance by the private sector on formal and informal sources of finance moved together: In those years when private rural firms drew in more bank loans, they also drew in more informal loans. The simple two-way correlation between the two series of data based on PSS2002 is 0.33. In the 1990s, the relationship became negative (—0.05); in those years when rural firms received fewer bank loans, they received more informal loans. This must have been a costly outcome for rural private firms. The drying up of bank loans drove up the costs of the informal loans.
Our third source of information concerns bank financing of fixed-asset investment activities. Fixed-asset investments - purchases of new equipment and property - are heavily financed by external capital. The high level of fixed-asset investments by rural households in the 1980s suggests the availability of external finance. Do we have direct evidence that this was the case? The answer is yes, although the information is not complete.11 According to Lin Senmu (1993), a senior official in the State Planning Commission, between 10 and 20 percent of fixed-asset investments of the individual economy were financed by bank loans in the mid-1980s. The NBS (1988, p. 560) provides data on bank loans for private fixed-asset investments in 1987. In that year, the total amount of bank loans for the private sector was 5.1 billion yuan, all of which was in the rural area. This was about 7.3 percent of the rural private fixed-asset investments.
We already saw in Chapter 1 that rural private fixed-asset investment was to fall sharply in the 1990s. Bank financing fell even more sharply. Throughout the 1990s, bank financing as a ratio of rural private fixed-asset
144
Capitalism with Chinese Characteristics
investment hovered between 3 and 4 percent, half of the level prevailing in 1987. In 2003, rural households invested 320 billion yuan. Of this amount, 12.5 billion yuan was financed by bank loans (NBS 2004b, p. 447), accounting for 3.9 percent of total investments. In 2004, bank financing fell further, to only 2.7 percent (NBS 2005c, p. 435). This is not even remotely close to the level prevailing in the 1980s.
3.2 Financial Liberalization in the 1980s
Some comrades asked about lending to individual business owners (getihu). For example, do you lend to him if he applies for $1,000 to import equipment? My opinion is that as long as his business is permitted by policy (fuhe zhengce) and contributes to economic development and as long as he has a permit from the Industry and Commerce Bureau and he can repay, of course, you can lend to him.
- Jin Deqin, President of the Bank of China, October 18, 1984.
Rural areas need state-owned banks and credit cooperatives for finance but at the same time, under bank supervision, we need to allow the existence of private (siren) free lending and borrowing.
- Han Lei, President of the ABC, July 20, 1984 (Han Lei 1984, p. 51)
The easing of the financing for rural private entrepreneurs did not occur by chance; it occurred because financial policy was tacitly or even explicitly supportive of the private sector. This is one of the least known aspects of the 1980s. Between 1980 and 1988, the Chinese financial system became increasingly flexible as the reformers directed banks and RCCs to lend to the emerging private sector. They also introduced proactive reforms of financial institutions by reducing state controls of RCCs and permitting entry by private players. The two quotes that begin this section are telling of this era. The first quote comes from the president of the Bank of China, Mr. Jin Deqin. In the 1980s, the Bank of China was tightly controlled by the central government as it was charged with the management of China's foreign exchange, considered a vital strategic and, at that time, scarce financial asset. Even in the area of foreign exchange, as early as 1984, officials were already expressing a willingness to lend to private entrepreneurs. This was only eight years after the end of the Cultural Revolution - a sort of financing re-engineering by comrades, if you will.
Second, the two men who made these statements were not some random financial officers; they were, respectively, president of the Bank of China and president of the Agricultural Bank of China. This underscores an important point about the 1980s: The financial reforms in the 1980s were not a stealthy
A Great Reversal
145
act by renegade local officials behaving against the controlling strictures by the central policy makers. The financial reforms were enacted by the central policy makers themselves. Later in this section, I provide statements by the topmost financial officer of the country - the governor of China's central bank - to illustrate this very point. The financial reforms in the 1980s did not occur randomly and haphazardly.
To be sure, these measures did not amount to a full-scale financial liberalization. Financial controls remained tight in the form of lending quotas and interest-rate caps, and the urban areas were immune to these financial reforms. Also, it is true that not all of the proposed measures were fully implemented. Keep in mind that the reformist leaders in the 1980s had only a few years to implement reforms, in contrast to the long tenure of the leaders in the 1990s (from 1989 to 2002). There were also policy reversals. For example, in 1986, to curtail the rapid credit growth, the ABC sharply curtailed credit supply to individual business owners.
These caveats aside, it is important to document and provide a paper trail of the rural financial reforms in the 1980s. A running theme of this book is that analyzing Chinese reforms is about ascertaining the directions of institutional or policy change, not about the level of institutions and policies. How to characterize the rural financial reforms is an art, not a science. One could argue that these were modest changes in making Chinese financial institutions more "business friendly" (in the sense that Hausmann, Pritchett, and Rodrik [2004] analyzed the policy changes in India under Rajiv Gandhi). Or one could argue that reducing the blockage of competitive entry into the financial sector and making credits available to the private sector marked a monumental change from the central planning era. My emphasis throughout this book is on directional liberalism as the most relevant benchmark. Regardless of one's views of the rural reforms in the 1980s, the fact is that rural financial practices were trending in a liberal direction in the 1980s and in an illiberal direction in the 1990s. Getting the China story right requires a dynamic perspective.
The pioneer in the financial reforms of the 1980s was unquestionably the ABC. This is not surprising given the fact that the rural reforms were at the forefront of the economic transformation in the 1980s. (In the 1990s, due to massive mismanagement and conservative reversions, the ABC became the most problematic bank in China.) In December 1984, the ABC unveiled its "Provisional methods of lending to industrial and commercial rural households" (Agricultural Bank of China 1986 < 1984>). In the same year, the ABC authorized floating interest rates for loans to individual business owners and waived loan-guarantee requirements for those borrowers
146
Capitalism with Chinese Characteristics
with a good credit history and with a high self-funding ratio (Agricultural Bank of China 1986 <1984>, p. 364). In 1988, after passage of the Private Enterprise Law, the ABC revised its 1984 regulations and added private rural enterprises - as opposed to the less ideologically sensitive household businesses - on the list of firms eligible for its non-farm loans (Agricultural Bank of China 1988a).
A consistent theme running through the bank-policy documents of the ABC in the 1980s is that the ABC and the RCCs should provide loans to rural residents to engage in non-farm activities. An ABC document dated July 1984 reveals that loans provided to finance commercial production by rural households increased between 30 and 50 percent "above the targets set for the year." The document, transmitted to all RCC branches in the country, describes the success story of a client of a RCC in Hunan province -clearly intended as an exemplary model for other RCC branches to follow. In this case, 28 farmers jointly founded a business, specializing in sourcing and distributing agricultural produce. Its operations were massive, sourcing from 17 townships and selling to 13 cities located in 5 provinces (Agricultural Bank of China 1984).
The most important financial institution in rural China was the rural credit cooperative. In 1985, RCCs accounted for 76.8 percent of all agricultural loans and 47.8 percent of all loans extended to TVEs. These numbers understate, however, the true importance of the RCCs. Many of the loans originating from the ABC were actually handled by the RCCs (China Finance Association 1986, p. 11-19). The RCCs were first established in 1951 as genuinely private financial institutions. RCC members elected the officers and determined the lending priorities and criteria of the RCCs in their respective regions. In the 1960s and 1970s, RCCs lost their operating autonomy and were placed under the administration of the ABC and local governments.
One of the first acts introduced by the reformist leaders was to move the RCCs back to the management system prevailing in the 1950s. This vision was mapped out in 1980, at the very start of rural reforms. In that year, the Politburo convened a finance leadership group specifically dedicated to the issue of rural financial reforms. The principle formulated by this group was called "restoring the original three features of the RCCs," meaning that they would be organizationally reliant on RCC members, managerially democratic, and operationally flexible. The 1980 policy document is remarkable in many ways. For one thing, it shows that in the 1980s, the financial reforms in rural China were being implemented at the same time as the general economic reforms. There was no lag in timing. Second, the document -issued only four years after the end of the Cultural Revolution - harshly
A Great Reversal
147
criticized the "government-run" nature of the RCCs. (The Chinese term for government-run is guanban.) This criticism of RCCs as a "government-run" institution appeared in numerous bank documents in the 1980s.
Starting in 1983, the Chinese state began to take concrete steps to implement this vision. Under the reform plan, the RCCs would put aside 30 percent of their deposits at the ABC as reserve and the RCCs would determine how to use the rest of their deposit capital on their own. The RCCs were also allowed to compete directly with banks both in the deposit-taking business as well as in the loan business (Agricultural Bank of China 1985 <1983>-b). By the end of 1985, 80 percent of the RCCs in the country had adopted reforms alongthese lines (Agricultural Bank of China 1986 <1985>,p. 34). Governance reforms of the RCCs began in the late 1980s as the RCCs moved toward more operating autonomy. In 1988, the ABC drafted a regulation on RCC employment practices. Article 11 of the regulation specifies that the local heads of the RCCs should be selected through "democratic elections" (Agricultural Bank of China 1988c, p. 200). The same regulation also discourages a common administrative practice in the Chinese bureaucracy -the rotating of heads of departments across geographic jurisdictions. My point here is not that the RCCs became truly self-governing credit cooperatives in the 1980s. Rather, my point is that in the 1980s, the RCCs were moving explicitly in the direction of autonomy and self-governance, whereas in the 1990s, the RCCs moved in the opposite direction, both in letter and in spirit.
Another sign of policy flexibility was the treatment of informal finance. In both the 1980s and 1990s, informal finance played an active role in meeting the financial needs of rural entrepreneurs and households. The difference, however, is that informal finance was not only tolerated in the 1980s but also was actually used by the reformers to benchmark the reforms of the formal financial institutions. In the 1990s, there was a protracted, costly, and ultimately futile effort to stamp out informal finance on the one hand and to intervene and micromanage the operations of the formal financial institutions on the other. The combination of these two led to substantial credit constraints in rural China in the 1990s.
The official stance toward informal finance in the 1980s was extraordinarily liberal. There were periodic crackdowns on specific private money houses (usually after fraud was discovered and a large amount of money was lost). But there was no attempt to stamp out the entire sector of informal finance, a huge difference with the 1990s, as I detail later. Again, the policy stance was not just an ad hoc, grudging official recognition of the actual informal finance practices on the ground, but rather it represented
148
Capitalism with Chinese Characteristics
a positive endorsement. The best way to illustrate this point is probably to point to statements made by none other than the governor of the PBoC. In a 1986 speech, Governor Chen gave a highly positive assessment of the financial experimentation in Wenzhou, the most liberal and the most capitalistic region in China. It is worth reproducing some of her statements at some length to illustrate the explicitly pro-private stance of the central bank at the time (Chen Muhua 1987):
• " [T] he gradual formation of Wenzhou's capital market is suited to the requirements of commercialized production in Wenzhou. In addition to the capital provided by the state banks and rural credit cooperatives, there are now various kinds of businesses with deposit-taking and lending operations. Non-governmental (minjian) capital mobilization and non-governmental rural cooperatives have emerged. The various methods of financial mobilization have made a positive contribution to local economic development."
• "Although the comrades working in banks, credit cooperatives, and insurance companies have made a lot of efforts to mobilize a substantial amount of capital and to support the legitimate financial requirements of economic construction, the needs of economic development are still not met. Now, there are so many non-governmental cooperatives in Wenzhou, with interest rates so high and with so much cash injection. There are so many rich people and so many speculative activities. All of these suggest that our banking work is not adequate, which calls for solving these problems by deepening the reforms of the financial system."
To be sure, Madame Chen did not give blanket endorsement to the various financial practices in Wenzhou. In particular, she singled out clandestine pyramid schemes known as escalating associations (taihui) for criticism. But the overall tone of her speech, as these excerpts show, was unmistakably positive. (She began her address to the Wenzhou government officials and bank managers by stating, "Today, I am not here to make a speech. I am here as a student.") She endorsed lending to private enterprises by the state-owned banking system, interest-rate flexibility, and the operation of private financial institutions subject to certain regulatory limitations. This high-level policy endorsement is particularly noteworthy considering the fact that, as Tsai (2002) reveals, there were some large-scale collapses of private financial houses in 1985 and 1986. (The money house run by Zheng Lefang, the Wenzhou housewife executed in 1991 for financial crimes, collapsed in 1986.)
A Great Reversal
149
Throughout the reform era, Wenzhou served as a barometer of the fundamental policy orientation toward the private sector. Madame Chen's highly positive assessment of Wenzhou exposes one of the biggest myths about the Chinese reforms - that the Chinese reforms were somehow pushed clandestinely by liberal local officials who connived against a conservative and controlling central leadership. Nothing can be further from the truth, at least for the 1980s. The private sector succeeded in Wenzhou because of the actively permissive, if not encouraging, stance of the central leadership in the 1980s. Madame Chen revealed in one of her speeches that the central government had sanctioned financial reforms in Wenzhou as early as 1982. The Chinese financial regulators had full knowledge of and endorsed many of the financial practices in Wenzhou. We know this because the president of the ABC, Dai Xianglong, who was to assume the governorship of the PBoC in 1995 and who cracked down on rural informal finance in the 1990s, detailed the practices of the Wenzhou RCCs in a 1987 speech.12 Ma Yongwei, a senior manager at the ABC, hailed the "new breakthroughs" by the RCCs in Wenzhou in moving toward flexible interest rates and achieving fund mobility across different regions (Ma Yongwei 1987 <1986>,p. 85).
Another noteworthy aspect of her speech and several bank documents from this era is the implicit, and sometimes explicit, view that the state-owned financial system was not competitive enough to satisfy the funding requirements in rural China. Chen's point that "we are not doing an adequate banking job" is entirely consistent with the main thrust of the RCC reforms - to decentralize the control rights of the RCCs so that they would be more responsive to the needs of rural households. In the long run, as the formal financial institutions became more competitive, this thinking goes, the market positions of loan sharks and usurious financial practices would be undermined. It was a remarkably market-based approach rather than an administrative instinct that sought to criminalize informal finance.
Because of the high degree of policy tolerance, informal finance flourished in many regions of the country. Western scholars believe that informal finance emerged mainly in the free-wheeling and dynamic regions of southern China, such as Wenzhou of Zhejiang province (Tsai 2002). A factor that correlates more strongly with the informal finance is not geography but the extent of private-sector development. Keep in mind that informal finance itself is a form of private entrepreneurship and its operations are both a result of and a condition for the flourishing of private businesses. Thus, one finds informal finance wherever private entrepreneurship was present and,
150
Capitalism with Chinese Characteristics
as I pointed out in the last chapter, private entrepreneurship thrived in the poor, rural, interior provinces. It is thus not surprising to find informal finance in those regions as well, not just in coastal provinces.
Take the example of Guizhou, a province that had a vibrant private sector in its rural areas. Guizhou, China's poorest, agricultural, and landlocked province, experienced surging informal finance activities in the 1980s. Streets were lined up with pawnshops and rotation associations (Editorial Committee of Guizhou Pan County Financial History, 1994), very similar to the description of the back-alley finance provided in Kellee Tsai's book about the more developed parts of the country. Informal finance was even present in a province known as the stronghold of the urban SOEs, Jilin province in China's northeast (home to one of the oldest and most established SOEs, First Automotive Works). A study conducted by the Jilin branch of the PBoC in 1987 reveals that 68.9 percent of the rural households in the survey borrowed from the informal credit market. The investigation details the uses of informal credit: 81 percent of underground loans were used for production purposes.
This study is revealing of both the extent of the informal finance in Jilin as well as the policy orientation of the PBoC in the 1980s. Jilin is not known as a pioneer in the economic reforms and has a well-deserved reputation of being cautious and economically conservative. Yet, PBoC's Jilin branch was highly positive in its assessment of the role of informal finance and concluded that informal finance "eliminated some of the inadequacies of the bank credit and contributed to the commercialization of the rural economy" (Jilin Branch of the People's Bank of China 1987, p. 151). That informal finance was present both in regions with an initially low endowment of state assets - such as Wenzhou - and in regions with a substantial presence of SOEs - such as Jilin - is evidence that the permissive stance toward informal finance was a central government policy rather than a discretionary policy of the local governments.
The best example of the market-based view of the Chinese reformers in the 1980s is the financial innovation called rural cooperative foundations. The background to the RCFs was the large-scale privatization of collective assets in the early 1980s. Although rural China privatized the control rights of collective land, some of the assets, such as plow animals or heavy-duty equipment, either were too expensive to be acquired by individual households or were indivisible assets - a donkey cannot be divided in two. So these assets still remained on the books of the villages, but they became illiquid as the collective entities shed their production role and the ability to generate revenue.
A Great Reversal
151
The RCFs rose in response to this problem. Villages securitized the collective assets by selling ownership shares to the members of the villages. The funds pooled from what amounted to private placements were then used to meet the short-term liquidity needs of the members of the villages. After the first round of initial privatization of the collective assets, the role of the RCFs migrated to something akin to the role of a savings and loan institution. The RCFs began to compete directly with official savings and loans institution such as the RCCs. In the 1980s, many RCFs were explicitly private, and in many ways they represented the model of what the financial reformers wanted the RCCs to become. (Some Western researchers believe that RCFs were tightly controlled by the local governments. They were, to some extent, but much of the research on RCFs was conducted in the 1990s, reflecting the state of affairs in that decade.)
The scale of the RCFs was enormous. By 1990, the RCFs covered more than 38 percent of Chinese rural townships (Rural Work Leadership Team of Fujian Communist Party Committee 1997). In 1990, the RCFs in Wenzhou pooled 20 million yuan from their members. This was a huge sum of money. In the same year, the total outstanding loans by the ABC amounted to 26.5 million yuan. At least in Wenzhou, by the end of the 1980s, the RCFs were beginning to approach the ABC in both size and reach (Editorial Committee of Wenzhou Financial History 1995, p. 152 and p. 225).
The RCFs are an excellent illustration of the fundamental differences between the 1980s and the 1990s. In the 1980s, policy makers wanted to make the RCCs more autonomous because they wanted the RCCs to become more competitive vis-a-vis other institutions such as the RCFs. Despite the fact that the RCFs competed with the RCCs, the Chinese government did not stamp out the RCFs. This was remarkable considering that the RCFs were never explicitly recognized by the PBoC as a legitimate financial institution. Unlike other financial institutions that were either regulated by the PBoC or were operating illegally, the RCFs enjoyed a semi-official status because they were loosely supervised by the Ministry of Agriculture, the most reformist central ministry in the 1980s.
3.3 The Financing Repression of the 1990s
The financing of the private sector contracted immediately after the Tiananmen crackdown. In 1989 and the 1990s, the credit financing of rural private fixed-asset investments amounted to half of the level in 1987 and 1988. Fixed-asset investments by the private sector slowed down substantially, as we saw in Chapter 1. During this period, the collective sector began to
152
Capitalism with Chinese Characteristics
receive the bulk of bank loans. Data show that Henan, a province that had a large private TVE sector, expanded loans to collective TVEs enormously. In 1984, the household-to-collective ratio of loans by RCCs was 1.90; in 1993, the ratio was 1.02.
The motivation in part was political in nature as the conservative central planners mounted an ideological assault on the private sector. The other reason was the macroeconomic retrenchment effected through a tightening of the credit supply. During the reform era, in years that inflation was high, private-sector development tended to be robust. For example, during the heyday of private-sector development in 1984 and 1985, rural credit expanded rapidly. In 1985, the ABC took in deposits of 93.4 billion yuan and lent out 168.5 billion yuan, injecting liquidity into the rural economy (People's Bank of China 1987, p. 4).
One of the few ways to finance the private sector, which operated outside the credit rationing plan, was to create more credit. This is why reformers such as Zhao Ziyang always seemed to have favored an expansive credit policy whereas the private sector tended to suffer under the inflation hawks, such as Li Peng. Credit extensions to the private sector are the micro-economic link between reforms and inflation. By implication, during the macroeconomic retrenchment, the private sector becomes the first policy casualty.13
The political assault on the private sector ended in 1993 after Deng Xiaoping conducted his famous "Southern Tour." The substantial financial repression of the private sector that occurred after 1993 was not motivated by political ideology but rather by technocratic ideology. The private sector, much of it rural, small-scale, low-tech, and hailing from the poorer parts of China, was considered not worthy of the country's precious financial capital. Much of the capital, then, was directed to what were considered the high-tech, urban parts of the country.
The financing repression of the private sector took two forms. One was a change in the lending priorities of Chinese banks. Banks were now instructed to support agriculture rather than to support rural entrepreneurs transitioning out of agriculture. This is an industrial policy mentality par excellence. Because of the view that agriculture is strategic - ensuring cheap agricultural supplies to industry and to cities - and because of distrust of the price mechanism, the idea is that the state had to use policy levers to affect the relative returns between farm and non-farm activities. Restrictions placed on non-farm activities were used to raise the relative returns on farm activities so as to ensure a steady supply of agricultural produce. The same rationale justified loan subsidies to agricultural production.
A Great Reversal
153
The other form of financing repression was a retreat from the financial opening and the tacitly encouraging stance toward competition that had prevailed in the 1980s. Private, unsanctioned financial intermediation competed with the financing tools of the state's industrial policy and reduced the ability of the state to direct resources. Thus, at an ever-intensifying pace, the state began to crack down on informal finance.
The RCCs, the primary credit facility to the non- farm rural entrepreneurs, were ordered to focus on agriculture. The ABC issued numerous specific quotas for the RCCs. Usually, 40 percent of the new lending was to be allocated to agricultural projects; in the agricultural provinces, the ratio was at an even higher level. Agricultural producers had priority over other borrowers, and those providing agricultural services, such as processing, transportation, and trading of agricultural produce, enjoyed top priority (State Council 1996). In the 1990s, the RCCs were used as a mechanism to subsidize agriculture. This policy development explains the drastic slowdown in non-farm business income in the 1990s. Because strictly agricultural activities always have a lower value added, the sectoral restrictions also explain the reduction in the overall growth of rural income.
Non-farm lending by RCCs was not banned but it was scaled back. Basically, non-farm lending became a residual. RCCs were to lend to non-agricultural projects only after the agricultural lending was fulfilled. This is how Dai Xianglong, the governor of the PBoC, outlined the priorities of the RCCs: "After the priority lending to satisfy the capital needs of agricultural production, if there are still funds available, then the capital needs of the TVEs and other industrial and commercial businesses can be considered" (Dai Xianglong 1997). The loan qualification requirements were made more stringent. The self-funding portion had to be 60 percent, an increase from the 30 to 50 percent range specified in the bank documents of the 1980s (People's Bank of China 1999, p. 146). Article 37 of the 1995 Loan Guarantee Law explicitly excludes plots of land for private farming and private housing as collateral assets.14
The restriction of RCCs to agricultural lending amounted to effective credit constraints on rural private entrepreneurs, the vast majority of whom started their businesses to get out of agriculture. This represented a fundamental shift from the focus of the RCCs in the 1980s, which was to facilitate the transition of rural residents out of agriculture. In Wenzhou, for example, a high share of RCC loans had gone to non-farm projects in the 1980s, about 39 percent cumulatively between 1984 and 1990 (Editorial Committee of Wenzhou Financial History 1995, p. 149). Another form of discrimination, more implicit than the sectoral restrictions, was that bank policy favored
154
Capitalism with Chinese Characteristics
production over construction of new facilities. In a 1996 State Council document, fixed-asset loans were capped at 30 percent of all RCC loans. This handicapped the private entrepreneurs, who had just started their businesses and needed capital to construct new facilities.
The discrimination against private entrepreneurs was not just de facto but was de jure as well. In the 1990s, there was a rising sentiment among Chinese financial regulators that private entrepreneurs posed higher credit risks (despite the mountains of evidence to the contrary). In 1992, the ABC (Agricultural Bank of China 1992a) instructed the RCCs to mandate individual business owners and private enterprises to deposit "a risk guarantee fund" before loan disbursements. Although private borrowers always faced higher costs, this policy was quite onerous. In the 1980s, the RCCs had charged higher interest rates to private borrowers as a way to mitigate the perceived risks of these borrowers. Good borrowers could generate profits to ease the higher interest costs. But the 1992 policy required an upfront payment and it made no distinction between a good and a bad borrower.
Chinese financial regulators felt that even this safeguard was insufficient. In 1994, the ABC issued another rule requiring the RCCs to impose an extra hurdle on loan approvals for individual business owners and private enterprises. Each loan to private entrepreneurs required two signatures, one from the loan officer and the other from the director of the regional RCC (Agricultural Bank of China 1994). This development is behind the observation by Park and Shen (2000) that loan approvals were centralized.
In the 1990s, the greatest change in rural finance was the increasing bureaucratization of the RCCs. Recall the 1988 draft regulation by the ABC to envision a system of selecting the RCC leadership on the basis of competitive elections by RCC members. This system was to replace the appointments of RCC managers by the ABC. This reform was now discontinued. Throughout the rest of the 1990s, among the large number of bank documents on RCCs issued by the ABC or the PBoC, not a single one refers to this 1988 draft regulation. Instead, management of RCCs was centralized. In a 1993 document entitled, "An opinion to speed up the rural financial reforms and opening," the ABC (Agricultural Bank of China 1992a) stated, "On the basis of the current leadership system, the emphasis should be on changing the operations and increasing the flexibility of the RCCs at the grass-roots level." The emphasis of this statement is on "current leadership system," signaling that the management system of the RCCs was not going to change.
Some of the key phrases used in connection with the RCC reforms in the 1980s disappeared in the 1990s. The three characteristics that the
A Great Reversal
155
reformers wanted to restore to the RCCs - that they be organizationally reliant on RCC members, managerially democratic, and operationally flexible - did not appear in any of the copious bank documents issued on RCCs in the 1990s. Other terms that did not survive in the 1990s include siren or siying, meaning private-run, or minjian, meaning non-government. In its stead, the operative term used to describe the RCCs was cooperative (hezuo). In Chinese parlance, these terms have very different and specific connotations. The term "cooperative" falls into the same category of terms such as collective (jiti). Cooperative and collective institutions usually have some private revenue rights but their control rights are effectively governmental. (For example, in the mid-1950s, the production cooperatives formed by the peasants were viewed as a transitional stage between private and state-owned means of production.) Minjian, siren, and siying all imply full private ownership defined as both private revenue and control rights. In this connection, the 1988 ABC decision to experiment with the democratic election of the RCC management is fully consistent with the vision to make the RCC a minjian institution.
Almost every other year in the 1990s, the State Council or the PBoC would issue a major decision on "accelerating and deepening reforms" in the financial sector. This is a salient feature of the financial-sector policies in the 1990s. A close reading of these "reform" measures almost always reveals in essence a centralization of control rights, an increase in the extent of micromanagement by the government, and a restriction on the activities of private actors in the financial sector. This is especially true of those measures concerning the operations of the RCCs.
One example is the sanctioning of rotating RCC branch directors in an ABC directive issued in October 1992 (Agricultural Bank of China 1992b). A rotation means that an official of one region is assigned to a position at the same bureaucratic rank in a different region. The practice is a mockery of democracy as it nullifies any election results in the rotated regions. (Imagine rotating the governor of Massachusetts to Maine.) For this reason, in the 1980s, the reformers explicitly discouraged the rotations of top managers of the RCCs. The practice was resurrected in 1992.
Rural finance became increasingly centralized in the 1990s. In 1996, after the authorities completely severed the administrative relationship between the ABC and the RCCs, the RCCs were placed under the administrative supervision of the local governments. In a top-down political system, this was a logical consequence. Control rights are always vertical, running from a higher level of the bureaucracy to a lower level of the bureaucracy. The PBoC actively encouraged the RCCs to link up with the local governments.
156
Capitalism with Chinese Characteristics
The specific mechanism was the control by the Communist Party. The idea of Party control of the RCCs was completely absent in all the documents related to the RCCs in the 1980s. It appeared for the first time in a 1998 policy speech by Shi Jiliang, the vice governor of the PBoC. He urged the county RCCs to be linked up (guaikao) and to actively report to the local Party committees (Shi Jiliang 1999 <1998>, p. 25). In another speech, Shi defined RCCs as "local government financial institutions," by which he meant that the local governments should exercise active leadership over the RCCs (Shi Jiliang 1999). These directives were explicitly contrarian to the letter and the spirit of RCC reforms in the 1980s.
Prior to 1996, the directives and rules issued by the ABC paid lip service to respecting the autonomy of the RCCs. After all, the RCCs were defined as "cooperative" financial institutions by the State Council in 1993. Toward the late 1990s, however, the PBoC dropped all pretenses, as indicated by Shi's definition of RCCs as local government financial institutions. Even the word "cooperative" appeared infrequently in bank documents. In March 1998, the PBoC issued a detailed decree, containing 35 articles, entitled "Provisional methods on managing the appointment qualifications of the principals of RCCs and union associations of RCCs" (People's Bank of China 1999 < 1998 >). According to the decree, the PBoC was to assume control over all aspects of personnel appointments, including the screening of candidates, account examinations (mandated after each principal's departure), and the termination of appointment.
By the late 1990s, after a decade of mismanagement, bad policies, and poor governance, the RCCs experienced massive operating problems and they contracted dramatically in number. In 1985, there were more than 400,000 RCCs in the country. This number was to decline sharply in the 1990s. In 1990, there were 384,320 RCCs and 286,389 in 1992. By 2003, only 91,393 RCC branches remained.15 This was the level of financial services in a massive country like China with more than 800 million rural residents. The performance of the RCCs also deteriorated. In the 1980s, the RCCs had non-performing loans (NPLs) on their balance sheets but most of them had been accumulated from the Cultural Revolution period. In 1994, 31.4 percent of the loan assets of the RCCs were non-performing and, in 1996, the NPLs increased to 38 percent, according to Dai Xianglong, the governor of the PBoC (Dai Xianglong 1997). The shareholder equity of the RCCs was reported to be 63.2 billion yuan in 1995, 54.8 billion yuan in 1996, 31 billion yuan in 1997, 15.1 billion yuan in 1998, and —8.5 billion yuan in 1999 (China Finance Association 1997, p. 452; 2000). In less than 10 years,
A Great Reversal
157
an institution that had contributed substantially to the takeoff of the rural sector was completely insolvent.
Despite - and most likely because of - the layers of detailed controls instituted by the government, the RCCs' lending practices became progressively more egregious. The following is a telling list in a PBoC document of the degeneration in lending practices of the RCCs (People's Bank of China 2001a):
• making loans to peasants in the form of goods rather than money and forcing peasants to sell the goods to designated buyers
• expropriating the share capital contributions of the members of the RCCs when extending them loans
• collecting taxes and fees from peasants when making loans to them
• making loans to township and village governments to finance their fiscal obligations to higher-level governments
• forcing peasants to purchase shares of the RCCs and deducting their share contributions from their loans
• building office buildings and purchasing sedans while operating at a loss
It is clear from this description that by the end of the 1990s, a decade of mismanagement and poor governance had made the RCCs into the policy pawns and cashiers of local governments. Corruption and fraudulent practices were rampant. The RCCs had completely stopped serving the financial needs of their members, a goal the reformist leaders in the 1980s had set out to achieve. The policy response to the mounting RCC governance problems is as telling of the policy makers in the 1990s as their measures that had created these problems in the first place: If centralization created performance problems, the solution was more centralization.
Instead of trying to resolve the deep-seated incentive distortions and increase transparency and accountability, the authorities opted for command and control at a rapidly escalating pace. In October 1995, the ABC issued the directive, "Provisional regulation on the auditing of rural credit cooperatives and punishment measures" (Agricultural Bank of China 1995). Regional branches of the ABC were to conduct regular audits of the RCCs within their jurisdictions and mete out penalties according to the provisions in the regulation. This was an exceedingly detailed decree containing 4 sections and 18 articles. All the penalties had a monetary price, ranging from 100 to 2,000 yuan. For example, the penalty associated with lending to an incorrect borrower, for the wrong uses of loans, for
158
Capitalism with Chinese Characteristics
loan contracts that do not meet specifications, or for borrowers violating loan contracts or government policy ranged from 20 to 1,000 yuan (Article 7). Provision 9 of Article 9 spells out the penalties for mishandling computer software, leading to data losses or the leaking of secrets. For some reason, mishandling computer software was considered more egregious than lending to the wrong clients, exacting a penalty starting at 100 yuan. Another rule issued by the PBoC concerned evaluations of the PBoC staff who monitored the RCCs (People's Bank of China 2001b). After all, those who monitor also need to be monitored. The style is classic command and control and the rule is numbingly detailed. (The set-up of a file system on the supervised RCCs is awarded two points in the evaluation, for example.)
In the 1990s, the authorities began to crack down on informal finance in a systematic fashion. The primary consideration was that informal finance was a source of competition with the state-owned financial institutions and that it drew resources away from the industrial policy programs of the state. The motivation was not the stability of the banking system. The available evidence indicates that financial institutions that were less tightly controlled by the state had better operating performance, but yet it was those state-owned institutions with poorer performance that were charged with the oversight of the better-performing institutions.16 The 1997 Asia financial crisis had very little to do with the crackdown on informal finance, which took place before the financial crisis and, in all likelihood, weakened China's financial system.
In the 1980s, the government tacitly tolerated the operations of the RCFs, but this policy stance was to change in the 1990s. In 1993, the State Council (State Council 1994 <1993>, p. 7) pointedly singled outthe RCFs, claiming they were not financial institutions and could not engage in deposit-taking operations. Instead, the role of the RCFs was to provide "mutual assistance" -small-scale, short-term revolving credit - to their members. The State Council decreed that the RCCs should take over those RCFs already engaged in deposit-taking businesses. In the next year, the government stepped up both the scale and the intensity of the campaign against the RCFs. The 1994 decision on restructuring RCFs prohibited lending and absorbing capital across different regions, and it established specific steps to absorb the RCFs into the RCC system. It also vastly limited the scope of the RCFs to agricultural lending because RCFs were not allowed to lend to urban residents or to develop branch networks beyond their home base (Rural Work Leadership Team of Fujian Communist Party Committee 1997).
The authorities stepped up the rhetoric against the RCFs in 1996. They were declared to be completely in violation of the financial regulations and
A Great Reversal
159
engaged in "vicious competition" with the state-owned banks for deposits (State Council 1996). The 1996 decree announced that all RCFs would be absorbed into RCCs, a change from the 1993 decision that absorbed only the deposit-taking RCFs into RCCs. The final blow to the RCFs came on July 13, 1998, when the State Council, in a decree signed by Zhu Rongji, categorically banned all informal financial institutions and practices, including RCFs (State Council 1998). The tone of the decree was extremely harsh, criminalizing all forms of informal financial practices and mandating the involvement of the public security bureaus in the investigation and punishment of the informal financiers. Those PBoC officials who failed to refer the cases to the public security bureaus were deemed to have committed criminal offenses (Article 27).
We began this section with a lengthy quote from the stern 1998 State Council order to close down, ban, or even prosecute the informal finance operations in the country. We also began this section with a 1984 quote from Governor Chen of the PBoC that sanctioned informal finance as a useful supplement to the operations of the formal finance. The two contrasting policy statements came from the very top decision makers - the 1998 State Council was signed by Premier Zhu Rongji - and they capture the essence of the difference in the financial policies of the two decades.
4 The Power of the Chinese State
The policy reversals on rural finance, financial reforms, and TVEs took place in a larger political context. From 1989 to 2002, China was led by a group of individuals imbued with heavy urban biases in their views of economic development and with a strong industrial policy conviction. In the 1990s, the key economic policy makers were all engineers by training.17 They followed a typical career path in a communist system - first serving as chief technicians and engineers at large SOEs and then ascending through the bureaucracy. Many of them came from overwhelmingly urban backgrounds. The top two national leaders in the 1990s, Jiang Zemin and Zhu Rongji, both had come from Shanghai prior to their elevation to their central posts. This represents a huge contrast with the 1980s when the top decision makers, such as Zhao Ziyang, Wan Li, and Tian Jiyun, gained prominence first as officials in the poor, agricultural provinces. (Interestingly, Hu Jintao comes from a background similar to the leaders of the 1980s, having first presided over Guizhou and Tibet in the 1980s.)
In the 1990s, FDI, technology, national champions, massive infrastruc-tural developments, and urban renewal were elevated to the top of the
160
Capitalism with Chinese Characteristics
economic policy agenda. In each one of these areas, the state is perceived as an indispensable instrument to make things happen. FDI was wooed through the construction of industrial parks and the bestowal of tax breaks. Technological acquisitions required state-sponsored and state-financed R&D programs. National champions were selected from incumbent large businesses, many of which were SOEs. Infrastructural developments and urban construction called for the intensive mobilization of a completely state-controlled resource in China - land assets. The direct economic role of the state in the 1990s remained substantial despite the fact that the Chinese state was shedding its ownership role.
In this section, I show that despite an economic transformation that is viewed by many as revolutionary, the size and the reach of the Chinese state have not diminished. In fact, by several measures, the Chinese state has grown massively since the early 1990s. Because Chinese capitalism is heavily rural in origin, the political environment in rural China has a direct bearing on private-sector development. Whereas governance deteriorated across the board in the 1990s, the extent of the deterioration probably was the most pronounced in rural China. One reason, apart from the economic policy reversals, was the strengthening of the political control by the state in China's vast countryside.
4.1 The Three Rural Crises
I have been a village cadre for nearly forty years. Even during the era of the commune system, control was never this tight. Today villages have no power whatsoever.
-A Hebei village official quoted in a research report published by the Development Research Center of the State Council (Zhao Shukai 2005)
In 2000, as foreign firms and Western analysts were celebrating China's prospects to join the World Trade Organization (WTO), inside China an entirely different sentiment prevailed. That sentiment is best captured by the term son nongweiji. San nongweiji- coined by Li Changping, a rural official in Hubei province, in his now famous 2000 open letter to then-premier Zhu Rongji - refers to the three rural crises: the crisis of agriculture, the crisis of village governance, and the crisis of the peasantry. Li details the egregious abuses of the Chinese peasants in the hands of local officials, the helplessness of the rural residents, and the onerous burdens shouldered by them. As Party secretary of a township in Hubei province and part of the political establishment, Li was in a position to know. He intimately understood the situation in the Chinese countryside.
A Great Reversal
161
Li's assessment contrasts sharply with outside views of China. In writings about the Chinese economy, there is a remarkable discrepancy between insiders and Western analysts. This is not because of a lack of information. In fact, in many cases, the Chinese government has been surprisingly and brutally honest. A 1996 report by the Politburo and the State Council acknowledges "severe incidents of clashes between cadres and masses leading to deaths and injuries." The report lists a set of banned practices, evidence that local officials were actively engaging in these practices, including dispatching the police to confiscate the money and property of the peasants and forcibly removing property and herds from the homes of the peasants (Rural Work Leadership Team of Fujian Communist Party Committee 1997).
I have already documented the recentralization of credit controls. In the 1990s, there was also a significant attempt to recentralize the administrative and fiscal management of Chinese villages. In the immediate aftermath of the rural reforms in the 1980s, there was a quick and initial decline in the power of the CCP Rural self-governance at the village level began to emerge. In the 1990s, however, there was an explicit and substantial effort to "rebuild" the CCP in rural China. Any progress that had been made in the direction of improving self-governance in rural China was eroded by the fiscal and administrative recentralization. This recentralization is captured in the statement - quoted at the beginning of this section - by a Hebei village official that "today's villages have no power whatsoever."
Very early on during the reform process, the CCP was already in a state of decline in rural China, a trend the Chinese state was determined to reverse in the 1990s. One immediate effect of the rural reforms was to render the title of Party secretary vacuous. The title did not connote any specific managerial or administrative responsibilities. In 1983, a village Party secretary issued what he called "a confession" - confessing to having nothing meaningful to do (Cui Anban 1983). An agenda of a township Party committee meeting contains rather mundane and marginal action items such as running an entertainment center for youth, conducting a campaign to extol courtesy, and cleaning the sidewalks ("A Report from the Shi Township" 1983). In his confession, the village Party secretary reveals that not a single person had applied to join the CCP for several years.
In the 1990s, the central government began to incorporate and then to increase the weight assigned to strengthening the local Party apparatus in its performance evaluations of subordinate officials. A major decision by the Politburo in 1994 laid out various measures to reclaim Party control of the countryside. The document sanctioned practices such as the stationing of higher-level cadres in villages and the appointment of outsiders to the
162
Capitalism with Chinese Characteristics
post of village Party secretary. A 1995 policy document prohibits private marketing and trade of fertilizers and reestablishes the state monopoly over grain procurement (Rural Work Leadership Team of Fujian Communist Party Committee 1997).
Based on survey research, academics Oi and Rozelle (2000) show that democratically elected village committees met infrequently, with the number of their meetings decreasing from 5.4 times per village in 1988 to 5.2 times in 1995. The low frequency of these meetings implies that truly important decisions were made elsewhere. Oi and Rozelle also report that Party members accounted for a very high share of the village representatives.18
A second related development was an administrative and financial recen-tralization of power in the hands of the townships that sharply curtailed the operating autonomy of villages. In the 1990s, the Chinese state resurrected some of the administrative practices that originated from the commune era of the 1960s and 1970s. For example, under the commune system, there was a practice called "area management" (guan pian zhidu), whereby designated township officials were put in charge of specific areas comprising several villages. The person in charge was called area head (pian zhang). In the 1990s, this practice was reinstitutionalized and vastly expanded. Even provincial officials were stationed in villages.
In the 1980s, China made tentative but meaningful efforts toward village self-governance. The Organic Law of Village Committees mandated popular elections of those village officials in charge of fiscal management, allocation of land rights, and education. In the 1990s, the maj ority of those sitting on the village committees were elected. But there may be a less benign explanation for this seeming success: The village committees had no real power. In the 1990s, the modest level of village self-governance was completely supplanted by the administrative and financial centralizations. The village elections were becoming increasingly meaningless because the township governments used the Party system to counteract the outcomes of the village elections. In a 2005 report, one township Party official was brutally honest when he said, "As for those village officials who do not obey the township Party committee and government, we can dismiss the [village] Party secretary. If we cannot dismiss the village head, then we can push him aside and not invite him to the meetings or use other ways to get rid of him" (Zhao Shukai 2005, p. 5).
This 2005 report reveals the extent of the micromanagement by township governments, rivaling that during the commune system period. Compensation for village officials is financed by the villages but the compensation norms and standards are set by higher-level authorities. There are four components to the compensation: (1) a basic salary, (2) a seniority component,
A Great Reversal
163
(3) a position salary, and (4) a discretionary component. The report is based on data from 10 provinces, as varied and diverse as poor provinces such as Ningxia and Gansu on the one hand and Zhejiang and Shandong on the other, but the monetary guidelines seem to follow a uniform standard across all provinces, another indication of centralization. For example, the seniority wage is set at 10 yuan per job-year and the position wage is set at 500 yuan for village Party secretary, 300 yuan for village head, and 200 yuan for village accountant. The discretionary component, implemented since 1996, is set by a detailed performance evaluation by the township, ranging from population planning, budgeting, Party building, law and order, tree planting, FDI, the building of schools, the paving of roads, and so on, by the township.
In the 1990s, villages lost independent budgeting power. According to a 2005 State Council report, written based on extensive field research, a practice called the "village account managed by the township" (cunzhang xiangguan) was implemented in the 1990s. The level of budgetary centralization is remarkable. The township governments set three approval levels -300 yuan, 500 yuan, and 1,000 yuan - above which the signatures of township officials, such as the deputy heads or heads of the townships, were required. The village cadre quoted at the beginning of this section was referring to this feature of township control. (He also said that he had enough authority to dig a small well.)
In a significant portion of the villages, the township governments not only approve budgetary applications but also directly take over management of the budget and cash-disbursement functions at the village level. This is called the "double centralized management" (xuan daiguan) - both budgetary approval and management at the township level. The 2005 report reveals that 16 villages had "single centralized management" and 14 villages had "double centralized management." Of the fifteen townships for which data are available, one started the budget centralization in 1991, five in 1997, two in 1998, five in 2002, and two in 2003. Thus, this was completely a 1990s phenomenon.
Let me underscore the huge operating implications of this move to transfer decision-making power from the villages to the townships. In 2005, there were 640,000 village committees and 18,900 townships. Moving decision making from the former to the latter entails a massive centralization of power. There is another issue here as well. In both the letter and the spirit of Chinese law, a village and a township are treated very differently in the Chinese political hierarchy. For example, the Organic Organization Law permits elections at the village level but not at the township level (Saich
164
Capitalism with Chinese Characteristics
2001). A village is not a formal part of the Chinese bureaucracy and village officials are not on the government payroll. Chinese political norms explicitly acknowledge the rights of villages to self-governance.
In contrast, a township is a formal part of the Chinese political hierarchy-it is the lowest level of the Chinese state. Township officials are officially on the government payroll and their expenditures are incorporated into the government budget. The township has an articulated government structure that is a near duplicate of the structure of the immediately superior government.19 There are also real ethnographic differences between a township and a village. A village is far smaller and, therefore, more close-knit than a township. In the 1980s, a village averaged around 30 households with a total of 150 people. Many Chinese villages are populated by members of the same clan. This is why these villages are known as natural villages. They have a cohesive and tight culture and kinship networks in a way that distinguishes them from large, artificial, and far more permeable townships.20 Centralizing the operating management of villages in the hands of townships nullified both the legal and the built-in autonomy of Chinese villages.
4.2 A Political Reversal
-"Individual laborers are the socialist laborer of our country.... As long as they meet the standards of the Party and [the Communist Youth] League, they should be recruited into the Party and the League according to the rules."
-From a circular by the Party Central Committee and the State Council issued on October 17, 1981 (Central Committee and State Council 1982 <1981>,p. 987)
"There exists, between private entrepreneurs and workers, a relationship of exploitation and being exploited. Private entrepreneurs cannot be admitted into the Party." -From a circular by the Party Central Committee issued on August 28, 1989 (Central Committee 1991 <1989>, p. 598)
"Since reforms and opening, the social structure of our country has changed substantially. There are now non-governmental high-tech entrepreneurs and technicians, managers and technicians employed at FIEs, individual households, private entrepreneurs, employees at intermediation organizations, and free-lance workers so on and so forth.... They are also contributors to the socialism with Chinese characteristics."
-From a speech given by Jiang Zemin on July 1, 2001 (Jiang Zemin 2006 <2001>, p. 286)
The last quote is from a famous speech Jiang Zemin gave on July 1, 2001. That speech is often described as path-breaking and credited as the one that finally conferred the belated political and ideological legitimacy on China's private sector. This view is simply incorrect.
A Great Reversal
165
The first quote is an excerpt from a circular issued by the Central Committee of the CCP and the State Council - in 1981. That circular already called for recruiting members of the private sector - called individual laborers at that time - into the Party. That circular addressed at great length and in great detail the need to equalize the economic and political treatment of the people working in the private sector with those in the state sector. The political environment for China's private sector started improving not in the 1990s but in the early 1980s.
Recall my account in the last chapter of Hu Yaobang's support for the emerging private sector in the 1980s. He coined the term "glory project" in 1983. This term was resurrected in 1994, by 10 private entrepreneurs, but there is a difference between how the term was used by Hu and how the term was interpreted in the 1990s. Hu Yaobang had argued that the economic contributions by the private sector were "glorious" but, in the 1990s, "glory projects" referred to the social contributions by the private sector - in the form of charity and donations to poverty alleviation and reforestation. An important feature of "glory projects" is noteworthy - it is specifically tailored to soliciting contributions from the private-sector entrepreneurs but not from the general corporate sector.
This is corporate social responsibility, Chinese style. Glory projects carry a rather subtle implication - that the charity contributions by the private sector are a form of indemnity against the political liabilities otherwise associated with private ownership. The unstated assumption is that the economic contributions by the private sector - output growth and employment generations - are insufficiently glorious. Private businesses need to make social contributions to make up their political deficit. This is very different from Hu Yaobang's original rationale for "glory projects."
This is a nice, if subtle, illustration of the substantial ideological hostility toward the private sector in the 1990s. Many assume that this ideological hostility was rooted in central planning and in the radicalism of the Cultural Revolution. This is true but part of the ideological hostility toward the private sector was revived by the leadership of the 1990s. The second quote is from a CCP document issued in August 1989 explicitly excluding private entrepreneurs from joining the Party.
I provide this documentary evidence here not to suggest that the ban was rigidly enforced in the 1990s but rather to argue that Jiang's 2001 speech eased the political and ideological restrictions that were created under his own leadership. It was the leadership of the 1980s that had taken on the ideological legacy of the central planning and Maoism; the leadership of the 1990s was revising its own views of capitalism. This is progress, to be sure, but let's give credit where credit is truly due. The easing of political
166
Capitalism with Chinese Characteristics
constraints in the early 1980s preceded and enabled the entry of private businesses. Despite the praise Western observers lavished on them, the achievement by the leaders of the 1990s amounted to providing the lagging political recognition of a private sector already substantial on the ground. Let's also keep in mind that the private-sector policies, as measured by fixed asset investments, became illiberal in the 1990s.
We do not know nearly as much about the politics of the 1990s as about the economics of that decade, but we can be certain of one thing - the Chinese state was not retreating. In the 1990s, the Chinese state reversed the gradualist political reforms undertaken by the leadership in the 1980s. This assessment comes from a well-placed insider, Mr. Wu Min, a professor at the Party School under the Shanxi Provincial Party Committee.21 In a 2007 article, Professor Wu reveals that the political reform program adopted at the Thirteenth Party Congress in 1987 made some substantial headway in terms of implementation during the one-year period after its adoption (a clear reference to the period leading up to the Tiananmen crisis in June 1989). According to Professor Wu, there were significant efforts to redefine and reduce the functions of the Communist Party. The Party committees were abolished in many government agencies and the functions of the Party and the state were explicitly delineated. Since 1989, however, despite the occasional rhetoric, there was no progress in the political reforms, especially in the area of reducing and streamlining the power of the Communist Party. Professor Wu argues that the stagnation of the political reforms is directly responsible for the multitude of the social ills plaguing China today.
The political reforms in the 1980s were designed to enhance the accountability of the government by creating some checks and balances over the power of the CCP and by fostering intraparty democracy. Professor Wu cites one specific measure in the 1990s to derail the reforms of the 1980s. According to Professor Wu, in the 1990s, China instituted explicit provisions prohibiting the National People's Congress (NPC) from conducting evaluations of officials in the executive branch, the courts, and the procu-ratorate. Professor Wu comments, "This is obviously a step backward and how can the system of people's congress be improved?"
Just how far did this step set back China? How about 1979? Three years after the end of the Cultural Revolution, the NPC began to exercise some real power. In 1979, in the aftermath of the capsizing of an oil rig during a storm in the Bohai Sea that resulted in 72 deaths, the NPC held hearings at which officials in the Ministry of Petroleum Industry were called to testify. The minister was determined to have been negligent and was sacked.22 (Incidentally, since the late 1990s, there have been numerous explosions
A Great Reversal
167
and industrial accidents in China's coal mines. Thousands of lives have been lost. Not a single official at the rank of minister or provincial governor has ever been held explicitly responsible.)
The stagnation or reversal of the political reforms was compounded by a substantial expansion of the scale of the Chinese state. Whereas the direct ownership role of the Chinese state declined, the magnitude of the state did not. In fact, since the early 1990s, the Chinese state has expanded in size substantially. There are several measures. One is a headcount of the number of civil servants. According to a researcher affiliated with the State Council, the number of officials on the government payroll was 46 million in 2004 (or 1 out of every 28 Chinese). In the early 1990s, the number was around 20 million. The researcher provides data on two poor provinces in China, Hebei and Anhui. In Hebei, between 1995 and 2003, the number of officials increased from 1.57 million to 2.19 million. In Anhui, between 1991 and 2003, the number increased from 1.2 million to 1.67 million (Zhao Shukai 2004b).
This expansion is especially noteworthy at the lowest level of the government apparatus. According to a 2004 government report, the number of township officials increased twofold during the course of the 1990s. In the mid-1980s, a small township had about 10 to 20 officials and a large township had between 20 and 30 officials. In 2004, an average township had more than 100 officials and sometimes even one department in a township had between 40 and 50 staff members (Zhao Shukai 2004a). The trend of these aggregate accounts is supported by micro survey data (although accounting differences mean that the match is not perfect). The fixed rural household surveys collected data on the number of village officials. In 1986, the number of officials per village was 6.2 persons; this number increased moderately to 6.29 in 1987 and 6.44 in 1988. In 1989, the number jumped to 9.08 persons. Between 1993 and 1998, the number of officials per village exceeded 7; it was 6.95 in 1999.
One relatively systematic measure of the size of the government is the fixed assets it has acquired for itself. Fixed assets here refer to the buildings, properties, and also possibly the vehicles operated by the government agencies.23 Along with the headcount of government officials, this is a superior measure of the size of the government than government revenues and expenditures. The fiscal size of the government is a better measure of the role of the government in the economy, not necessarily its size. (For example, the United States can have a large government budget relative to its GDP, but the employment size of the government itself is relatively small compared with the size of the private sector.)
168
Capitalism with Chinese Characteristics
Another advantage of the fixed-asset investment data is that they are organized on a systematic basis and the series go back to the early 1980s, which allows for an analysis of the trends over time. There is also more disclosure. Many of the operations of the Chinese government are shrouded in secrecy, but we have more information on the fixed-asset investment activities.
In 2002, the fixed-asset investments by the apparatus of the state - defined as the agencies of the government and of the CCP - were 137 billion yuan, or about US$17 billion by the exchange-rate conversion.24 This figure reflects the fixed-asset investments made by the entities of the state - government agencies and CCP departments. In the same year, the same state entities spent 56.6 billion yuan in fixed assets in the agricultural sector. In Chapter 5 of this book, I provide further evidence that the Chinese state today is self-serving. Here is a concrete illustration of this judgment: 800 million Chinese peasants claimed less than half of what 46 million Chinese bureaucrats claimed in fixed-asset investment resources. In the same year, the educational sector received 95.2 billion in fixed-asset investments, 68 percent of what the apparatus of the state invested in itself.
Let us also look at the trends over time. In 2002, the fixed-asset investments in the state apparatus amounted to about 7.1 percent of the total fixed-asset investments made by the state sector. In order to match this ratio, we have to go back to 1982 when the ratio was 7.0 percent. Here, once again, we have a tale of two decades. Throughout the 1980s, this ratio steadily declined, from 3.5 percent in 1985, 2.9 percent in 1988, to 2.3 percent in 1990. Beginning in 1991, the trend began to reverse. The ratio was 2.6 percent in 1991, 4.7 percent in 1995, and then 6.2 percent in 1998. By 2002, the ratio at 7.1 percent was more than twice the ratio in the last year of the 1980s.
4.3 An Industrial Policy State
The prevailing view in the West is that the Chinese state carried out a massive privatization program in the 1990s. An explicit turning point in the policy stance toward privatization is believed to be the 15th Party Congress convened in September 1997. Privatization did increase in scope and in intensity at that time when, by various estimates, between 30 million and 40 million workers were laid off from the SOEs (Garnaut, Song, Tenev, and Yao 2005; Yusuf, Nabeshima, and Perkins 2006).
Often missing in these accounts is another development that occurred during the same period: Massive investments in those large incumbent
A Great Reversal
169
enterprises in which the state retained substantial and controlling equity shares. This is the industrial policy aspect of the Chinese state in the 1990s. The Chinese government is explicit and completely open about its own industrial-policy proclivities. The policy program officially sanctioned by the 15th Party Congress was "grasping the large and letting go of the small." "Letting go of the small" was the privatization component of the program with which Western academics are familiar; "grasping the large" was the industrial-policy component seldom emphasized in the works on this period.
The industrial policy agenda shaped China's privatization agenda. The purpose of "letting go of the small" was to limit the scale of privatization. According to a government estimate, small SOEs accounted for only some 18 percent of the assets in the state sector as of 1997. However, small SOEs accounted for the majority of the losses in the state sector (State Development and Planning Commission 1998). This is mainly because the small SOEs had to compete with the non-state firms, whereas the larger SOEs were protected from competition. Many of them were monopolies.
The standard economic rationale argues that the state should privatize the profitable SOEs first. The idea is that the profitable SOEs can be privatized with minimum social consequences. They have fewer employment redundancies and presumably they can fetch higher bids because of their sound financial conditions. The privatization proceeds can then be used to ease the pains to pay for the social costs of restructuring and privatization of the unprofitable SOEs in the future (Roland 2000, p. 248). This is an impeccable rationale. Presumably, private investors and entrepreneurs are better at managing and growing assets to create economic value, whereas the state has a comparative advantage in managing social responsibilities.
In the 1990s, the Chinese government did exactly the opposite, putting the country through an unnecessarily socially wrenching process. The privatization of small SOEs meant the loss-making SOEs were privatized. This policy stance maximized the social costs while it minimized the economic benefits. The privatization program financed a substantial build-up of the capital of those large SOEs that the state chose to retain. With an increasing intensity and level of specificity, several policy initiatives - in 1989, 1991, 1995, and 1997 - all aimed at supporting or creating ever larger SOEs. In 1991, the government selected 55 enterprise groups for experimentation and, in 1995, it expanded the list to 57. In 1997, the list was expanded again, to 120 (Institute of Industrial Economics 2000). Most of the beneficiaries of the government's industrial-policy program have been the SOEs. The target firms were given tax and debt relief, import licenses, greater access
170
Capitalism with Chinese Characteristics
to domestic and overseas listing facilities, and substantially increased operating powers, such as powers to purchase and sell assets and to transfer assets across geographic and bureaucratic jurisdictions. The economic sectors covered by these firms also expanded to encompass virtually the entire economy. In 1990, the State Council issued a policy of "two guarantees" for 234 SOEs, guaranteeing them access to bank loans and raw materials. In 1994, the central government declared electronics, automobiles, petrochemicals, and construction to be the four "pillar industries" of the economy. The SOEs are dominant players in all of these four industries. Most of the 120 large enterprise groups slated for preferential policy treatments in the 1997 initiative were SOEs as well.25
The industrial-policy rationale for the SOE reforms forms another contrast with the 1980s. As is well known, Zhao Ziyang had begun to advocate SOE reforms in the late 1970s when he was Party secretary of Sichuan province, and Sichuan implemented one of the first programs to reform the SOEs.26 What is noteworthy is that Zhao advocated reforms of the SOEs when the SOEs were making huge accounting profits. Zhao and his advisers clearly believed that the SOEs, as SOEs rather than as loss-making businesses, lacked competitiveness. A policy that limits privatization only to loss-making SOEs is based on a view that SOEs themselves were not plagued by distorted incentives and political control problems endemic of the state ownership. Rather, the diagnosis is that SOEs incur losses because they lack resources, technology, and investment opportunities.
The approach of Zhao toward SOE reform focused on solving a control-right problem. His contract approach, at least in terms of design if not in terms of the actual outcome, was trying to get at this control problem. Under this approach, SOE managers would sign contracts with the state that specified the obligations to the state and assigned the residual rights to the managers. There are differences of opinion as to whether the reform was successful, but the specific outcome of the reform need not detain us here. The important point is that the contract reform reveals an underlying, if implicit, intellectual framework that identified the core problem of the SOEs - their political control rights. Zhao's approach did not work because of the lack of complementary reforms and because of the short time frame of his leadership.
A view focusing on the control-right problems of the SOEs ought to have led to the next logical step of contract reforms - management buyouts of the SOEs. But, in the early 1990s, the Chinese leaders reversed the policy on the grounds that the contract reforms did not work. Instead, they embraced an industrial policy approach that actually augmented the control rights
A Great Reversal
171
of those SOEs that the government had decided to retain. In the 1980s, collective TVEs, such as Kelon, had state revenue rights but private control rights. In the 1990s, in the case of the large SOEs, the situation was completely reversed. Most of the large SOEs, which were listed on China's two stock exchanges, had partial private revenue rights but complete state control rights.
Between 1990 and 2003, only 6.97 percent of the initial public offerings on the two Chinese stock exchanges were from private-sector companies. The rest were SOEs that issued minority shares but in which managerial control remained very clearly in state hands.27 Put differently, because many shareholding firms in China have private revenue rights but their control rights still rest with the government, they should be considered as state-controlled. According to a detailed study of more than 600 firms on the Shanghai Stock Exchange (SHSE) and the Shenzhen Stock Exchange (SZSE) in 1995, the three main groups of shareholders - state, legal persons, and individual shareholders - each controlled about 30 percent of the outstanding shares (Xu and Wang 1997). This stock split has remained more or less constant since then, although the government has plans to reduce the state shares. The control rights of these firms were overwhelmingly state. According to the same study, although individual shareholding constituted 30 percent of the outstanding shares, on average individual shareholders occupied less than 0.3 percent of the seats on the boards of 154 companies, whereas on average the state was over-represented on the boards. On average, the state retained 50 percent of the seats even though its equity shares amounted to 30 percent. There were no proxy voting procedures, thereby putting the individual shareholders in a disadvantageous position vis-á-vis the institutional investors such as the government agencies. This usurpation of rightful shareholder power is direct evidence that the state harbors no intention of relinquishing its control rights even over those firms that have explicitly private revenue rights.
5 Conclusion
A widely accepted paradigm to explain Chinese reforms is the gradualist perspective - the idea that the Chinese reforms deepened over time in an incremental fashion. The economic and political logic of gradualism is powerful.28 Gradual or incremental reforms build both political and economic complementarities. Reforms are fraught with uncertainties about eventual outcomes, and the best reform program minimizes these political and social costs and generates a bottom-up demand for deeper reforms because
172
Capitalism with Chinese Characteristics
the public and government can take advantage of the successes of the initial easy reforms. The economic logic is just as strong. In the case of China, for example, the entry of non-state firms reduced the SOEs' profitability, which forced the SOEs to reform. Naughton explains the feasibility of this self-enforcing reform mechanism in terms of the "interconnectedness" of the institutional features of the centrally planned economies. Reforms are contagious because "unhooking a single key connection can cause the entire fabric to unravel" (Naughton 1996, p. 311).
In this chapter, I question the claim that China followed a gradualist strategy in the 1990s. My argument is that the gradualist perspective fits the China of the 1980s but not the China of the 1990s. Many of the productive reforms in the 1980s were partially or completely reversed in the 1990s. Fiscal decentralization, which is credited as an important positive incentive for growth, was largely reversed in 1994. The control rights of the small SOEs that had been delegated to managers in the 1980s were recentralized in the early 1990s, although many were fully privatized in the late 1990s. Private-sector financing became more difficult in the rural area. The political reforms stagnated completely. By far, the greatest reversal occurred in rural China. The financial innovations to lend to rural households to start non-farm businesses and to allow private financial intermediation were discontinued.
These reversals imply real consequences. In Chapter 5, I show that the ratio of investment to Chinese GDP rose steadily in the 1990s, unlike other East Asian economies in which the investment/GDP ratio declined as they became richer. By 2005, China was investing close to 50 percent of its GDP, a level that we do not see anywhere else in East Asia. One possible explanation behind this rise of investment is a shift of sources of growth. Because of the increasing policy and credit obstacles placed on the indigenous private sector, the ability of entrepreneurs to contribute to economic growth by product and process innovations was suppressed. The repression of the broad-based, small-scale private entrepreneurship would also depress income growth, thus limiting domestic consumption as a driver of growth. To maintain the same pace of GDP growth would require increasing the investment levels. This hypothesis dovetails with the fact that China launched huge infrastructural and urbanization projects since the mid-1990s. A large portion of those investments occurred within the state sector.
The massive investment boom, however, happened at a price. In the concluding chapter, I come back to this issue and ask the question, "If China invested so heavily in transportation and urban infrastructures and
A Great Reversal
173
skyscrapers, what is it that the country is not investing in?" The answer, as it turned out, is education, especially education in the rural areas. In the 1990s, as China succeeded in creating world-class infrastructures, the government was charging fees for basic education and even for administering immunization shots to rural children. The result was a sharp rise of illiteracy and a slowdown in the pace of reducing infant mortality.
Another price of this investment boom is productivity growth. Beginning in the late 1990s, by some estimates, productivity growth slowed down and by other estimates it completely disappeared. (More data is presented in Chapter 5.) This is a very worrisome development. We know from economic research on East Asia that productivity slowdowns presaged the general economic contractions or even financial crises (Young 1995; Krugman 1994). China is facing a governance crisis of a significant magnitude. Corruption is rampant, the nature and scale of which began to take the form of grand theft in the 1990s, as opposed to the controllable, low-intensity corruption of the 1980s. The extreme policies, accelerating in intensity since the late 1990s -such as the laying off of tens of millions of workers without adequate social protection, the charging of ever higher fees for basic government services, the land grabs, and the growing crony capitalism - aggravated social tensions and contributed to income inequity. Social protests, some extraordinarily violent, are occurring at an increasing frequency.
I have already shown that the income growth of rural households slowed down dramatically in the 1990s. The TVEs, which were largely private, began to languish. I go into greater detail about the costs of the strategy in the 1990s in the final chapter of this book, but let me highlight a few of them here. One is that GDP growth in the 1990s increasingly was disconnected from the welfare of Chinese citizens. The ratio of household income per capita - gathered through surveys - relative to GDP per capita declined continuously during the decade. Yes, GDP was still growing rapidly, but each increment in the GDP growth entailed smaller improvements in the welfare of Chinese citizens. In the 1990s, education and health care were made more expensive and less accessible in rural China.
The true China miracle is a story of the 1980s when a vibrant rural entrepreneurial class emerged. This was the phase of what Baumol, Litan, and Shramm (2007) describe as entrepreneurial capitalism. The story of the 1990s is one of substantial urban biases, huge investments in state-allied businesses, courting of FDI by restricting indigenous capitalists, and subsidizing the cosmetically impressive urban boom by taxing the poorest segments of the population. This period is closer to what Baumol, Litan, and Shramm term as state-led capitalism. The epitome of this statist form
174
Capitalism with Chinese Characteristics
of capitalism took place in Shanghai, a city that was left untouched by the reforms in the 1980s but became a political power base in the 1990s. The Shanghai model, with its skyscrapers and Maglev train, has impressed countless foreign observers of China and has inspired both the admiration and the fear of a rising China. But, a hard look into Shanghai leads to a very different perspective. At its core, Shanghai is substantially state-controlled and state-led. Its private sector is very under-developed. Personal income has not grown nearly as fast as the GDP of the city. This is the subject of the next chapter, "What is wrong with Shanghai?"