5 / Making Xinjiang Safe for the Han?
Contradictions and Ironies of Chinese Governance in China’s Northwest
In 1949, when the troops of the People’s Liberation Army (PLA) entered Xinjiang, less than 10 percent of the population was Han. Historically, Chinese control over Xinjiang was sporadic, and more or less continuous control came only during the Qing dynasty (1644–1911). Xinjiang did not obtain provincial-level status until 1884 (not coincidentally, the same year Taiwan did). This occurred after Zuo Zongtang led a military expedition to Xinjiang to put down the Yakub Beg–led rebellion and reassert limited central control. In the Republican period (1911–1949), Han nominally continued to lead the province, but the central government’s reach was limited, and Xinjiang was arguably the least-integrated province in China under the weak Republican regime. Soviet “diplomats” provided support to rebel movements and various local potentates. So when the People’s Republic of China (PRC) was established in 1949, the connections between the Chinese heartland and Xinjiang were extremely weak.
Josef Stalin did not let Mao Zedong forget how weak those ties were. In a secret protocol to the Valentine’s Day (1950) Treaty of Friendship, Alliance, and Mutual Assistance between the PRC and the Soviet Union, both countries pledged that “citizens of third countries [would not be allowed] to settle or to carry out any industrial, financial, trade, or other related activities in Manchuria and Xinjiang.” Similar restrictions were to apply in the Soviet Far East and Central Asia.1 Soviet-Chinese joint stock companies were soon set up to explore for uranium and other rare-earth metals useful to the Soviet nuclear-development program, as the Soviet leadership thought that it lacked abundant uranium resources.2 The agreement struck the Chinese leadership as yet another unequal treaty, imposed on China this time by an ostensible ally. In addition to giving the Soviet Union special economic privileges in China, the treaty allowed the Soviet Union to maintain a number of consulates in Xinjiang; these consulates had ties with the non-Han population.3
The realization that integrating Xinjiang more fully into China proper was crucial to national security was reinforced by several circumstances and events: the 1962 migration of more than sixty-two thousand minority people into the Soviet Union in response to discrimination against minorities and Islam in the period from 1958 to 1962, the existence of road access to Tibet via Xinjiang, the Sino-Soviet battles of 1969, and the scattered military confrontations of the 1970s. The migration, or flight, of the sixty-two thousand minority people across the border, which came, according to the Chinese version, at the instigation of Soviet diplomatic personnel, reinforced traditional Chinese concerns regarding the linkage between external threats and internal disorder.
For much of its history, the PRC regime has been highly concerned with integrating Xinjiang economically, militarily, and politically into a modernizing Chinese state. Xinjiang has been the site of extensive investment by the center. But the economic activities of the Chinese state and its representatives in Xinjiang have not been ethnically neutral. Rather, they have overwhelmingly benefited the Han in Xinjiang, and many seem deliberately designed to encourage Han immigration. Today, the military presence in Xinjiang is extensive, and the overwhelmingly Han-dominated Production and Construction Corps (PCC) is an empire almost to itself, exercising production, military, paramilitary, and “judicial” functions. This unusual organization, which was composed initially of the Communist troops who occupied Xinjiang in 1950, has the equivalent of provincial status in Chinese economic planning and is not under the control of the Xinjiang Uygur Autonomous Region (XUAR) authorities. Chinese agricultural policies in Xinjiang appear to be creating an environmental disaster as the short-term need to encourage migration comes at the expense of sustainable development (if such a thing is possible in a region where no urban area receives more than ten inches of rain a year). It should come as no surprise to anyone that policies designed to increase the Han population of Xinjiang—policies that include offering Han economic incentives and reserving well-paying jobs for them—have solidified minority ethnic identities and increased anti-Han feelings (especially in the reform period from about 1978). Such feelings, which arose periodically even before Xinjiang was incorporated into China, have stimulated oppositional activities and led to a classic action-reaction pattern of violence.
This chapter focuses on a number of the economic aspects of what can only be seen as Han economic imperialism in Xinjiang. In comparison to other parts of northwest China and other nominally autonomous provincial-level units, Xinjiang has done quite well economically. The Han, however, have been the major beneficiaries of this process, and the trend appears if anything to be growing stronger. But economic colonialism (or internal colonialism) in Xinjiang can be understood only as a response to Han–perceived security threats, both internal and external to Xinjiang. This chapter also examines the role of the PCC in exerting control over and developing Xinjiang. The PRC has invested so much in, and so much depends on, Xinjiang that the regime will do what it must to retain undisputed control over this vast region.
FOREIGN POLICY AND XINJIANG SECURITY POLICY
Comprising a sixth of all Chinese territory and bordering on eight foreign countries (five, prior to the collapse of the Soviet Union), and with a longer border with foreign countries than any other provincial-level unit in China, Xinjiang poses security issues that have always been a major concern of the central government in China. The Xinjiang Production and Construction Corps was established in the early 1950s to help open up land and provide paramilitary and/or reserve-military support for regular military units in Xinjiang. At least in its early days, the PCC was composed overwhelmingly of demobilized soldiers, both from the PLA and from Guomingdang units that had surrendered in Xinjiang. It appears that Xinjiang was relatively pacific through mid-1957, though campaigns against counterrevolutionaries and other “bad elements” took place.4 With the Antirightist Campaign of 1957, policies of accommodation toward national minorities began to give way to campaigns that attacked local nationalism. Although these attacks would ebb and flow with general policies toward class struggle in the Chinese heartland, for the next twenty years there were regular attacks, and sometimes extended campaigns, against the leadership and culture of the non-Han peoples of Xinjiang.
The deterioration of Sino-Soviet relations, coupled with the flight of 62,000 people from northern Xinjiang in 1962, led increasingly to the militarization of Xinjiang’s extensive borders with the Soviet Union and Mongolia. Wang Enmao, then the leader of Xinjiang, desired moderate policies that would prevent ethnic conflicts from emerging and make managing Xinjiang’s increasingly serious external-security concerns easier. Despite Wang’s efforts to limit the effects of the Cultural Revolution, widespread violence and near chaos were at times common in Xinjiang. Indeed, though it is hard to say where the Cultural Revolution was the most violent or the most disruptive, it appears that in many, if not all, of the PRC’s “autonomous regions,” policies aimed against minority individuals and minority members of the Chinese Communist Party (CCP) were severe.5 As a result of the central government’s Third Front policies, Xinjiang militarized, and nine ordinance factories were built in the region, including a major one in Urumqi, north of the Tianshan mountain range, and another major one in Korla, south of the mountains.6 The ordinance industries of more populous border provinces such as Guangxi, Yunnan, and Inner Mongolia do not appear to have undergone such extensive development.7 By comparison, the ordinance enterprises in Xinjiang may have made up 10 percent of all the medium- and large-scale enterprises in Xinjiang in 1985.
Xinjiang was slow to recover from the Cultural Revolution, and there were reports of widespread conflict in the early post-Mao period. Ethnic conflict was reported in Aqsu, a border clash with the Soviets occurred in Tacheng, and agitation among youth sent down to the countryside who were anxious to return to their native places was widespread.8 This slow recovery also seems to have characterized the situation in many of the PRC’s autonomous regions. With the reform period after 1978 (perhaps best symbolized by Hu Yaobang’s visit to Tibet in 1980), the central government appears to have presented a “new deal” to the autonomous regions: it authorized some accommodations to minority culture, central subsidies to all autonomous regions appear to have increased, and as with the rest of the country, economic development became the priority task. It helped Xinjiang that beginning in 1982, China and the Soviet Union gradually began to improve their relations. The external-security threat diminished, and Han leaders tried to improve the internal-security environment.
During the 1980s, extensive efforts were made in Xinjiang to compensate for prior failings. From 1982 through 1996, non-Han students constituted a majority in Xinjiang colleges (though their percentage never matched their numbers in the general population). Central-government subsidies constituted one-half to three-quarters of all spending in Xinjiang. Greater efforts were made to create a better balance between light and heavy industry. Between 1979 and 1991, the ratio of heavy industry to light industry fell from 59–41 to 50–50, implying a greater concern for producing consumer goods and raising popular living standards. From 1980 to 1993, the overall membership of the PCC remained constant, while its Han membership declined marginally from 90 percent of the total to 88 percent.9 Reports of unrest in Xinjiang declined, and little information is available about social disorder until 1990 or so. Thus, the combination of improving relations with the Soviet Union and the granting of greater scope for minority people’s religious and cultural beliefs and practices appears to have improved the security situation in and around the region.
But this improvement in the security situation was gradually undermined by a variety of factors, many of which were extrinsic to Xinjiang. First, China’s demand for petroleum began to exceed its internal supply around the mid-1980s. China’s established fields in the east were not increasing production, there were few on-shore untapped reserves or promising geological structures in eastern China, and hopes for off-shore development were not matched by actual, exploitable finds. Xinjiang became the major locus of assumed on-shore potential. As a result, resources began to flow into Xinjiang in significant quantities in the mid- to late 1980s. Not only were prospecting and exploration carried out, but the heavy equipment needed for oil development called for enhancement of the infrastructure. On-shore oil development was at that time under the Ministry of Petroleum (now the China National Oil Corporation), a central-government organization that was dominated by Han. The new exploration teams in Xinjiang increased the Han population, and their wages were relatively high. Xinjiang’s potential oil thus became China’s last hope for petroleum self-reliance and an essential component of national economic security. But though there has been substantial expansion of Xinjiang’s production, huge new finds have not occurred.10
Second, the reemergence of Tibetan opposition to Chinese rule, which culminated in the protests of late 1987 and early 1989, prompted conservatives in the central leadership to question whether allowing somewhat greater latitude for differing cultural practices and identities was good policy, not just in Tibet, but elsewhere. Militant Islam, especially as seen in Afghanistan after the Soviet withdrawal, was also viewed as threatening. The 4 June 1989 crackdown on student protesters in Tiananmen Square dealt a crippling blow to all approaches for dealing with social and political issues in China that were not hard-line.
Finally, the collapse of the Soviet Union and communism in Europe further heightened the fear of Chinese political elites for future developments. Was communism doomed to failure? What had led the Soviet Union to collapse? Was a similar collapse in China inevitable? Was the fact that the Soviet Union was widely perceived as an empire, with a rapidly growing population of non-Russians, a major cause of its collapse? These questions reverberated throughout the leadership from mid-1989 until mid-1992.
Although the Soviet Union had been poised between 1969 and the early to mid-1980s to invade Xinjiang (in a way that would have been almost impossible to stop until the Soviet Army reached Gansu—if then!), the very existence of the Soviet Union (and Sino-Soviet tensions) kept the borders more or less closed and prevented strong cross-border links from developing between the Turkic peoples of Xinjiang and those of Soviet Central Asia. The Chinese leadership was unsure whether the Soviet Union’s successor states would abet “local nationalisms” within Xinjiang and whether the independence of Central Asia would echo among those believing in an independent Xinjiang or East Turkistan (or any other name).
China’s energy needs, the reemergence (from the center’s point of view) of restive minorities, and the collapse of the Soviet Union raised new security concerns in Beijing. The central government responded to these concerns in ways that were both innovative (in its external diplomacy and international economic relations) and reactionary (in its handling of internal security concerns within Xinjiang). It also adopted new economic policies in Xinjiang that greatly increased Xinjiang’s integration into Han China.
China moved quickly to recognize the new states of Central Asia and to initiate economic and other relations with them. In short, China attempted to co-opt them into contributing to China’s economic growth and political stability. Within three years, China, Russia, Kazakhstan, Tajikistan, and Kirghizstan had initiated a process of annual consultation, border delimitation, and mutual cooperation. The acceptance of and support for this “G-5” process was an unprecedented development in Chinese foreign policy. This cooperation has led to significant gains for Chinese security: the PRC has extradition treaties with the states of Central Asia, which have returned independent activists to China (where they were probably executed); it has obtained the agreement of these states not to support independence movements in Xinjiang; and it may have secured the right to engage in “hot pursuit” across international borders to destroy guerrillas who are working for independence or true autonomy for Xinjiang. The Chinese leadership can and should be well pleased with how its diplomacy has limited transnational political linkages between the Turkic peoples in Xinjiang and those across the border. This has given the agents of coercion in Xinjiang a generally free hand to employ extreme measures to try to suppress separatist trends; they have had mixed success.
The independence of Central Asia provided China with another opportunity to deal with its energy problems. The China National Oil Corporation successfully bid on the opportunity to develop a major oil field in western Kazakhstan and to build a thousand-mile pipeline to bring the oil from Kazakhstan to Xinjiang, where it would link up with an expanding pipeline network feeding into the rest of the country. Given the cost of the project (about $10 billion), one can assume that the central government stood firmly behind this deal. Actual progress on the oil field and pipeline has been slow, though Kazakh-produced petroleum is entering China by rail and truck. Provided the project actually gets built, access to the output of the large reserves in western Kazakhstan will have a major positive effect on China’s petroleum supply. But it may also increase China’s security fears. Pipelines are highly vulnerable, and they obviously have fixed paths and locations. A thousand-mile pipeline across Central Asia makes a good target for oppositional violence to Chinese rule in Xinjiang (whether the violence takes place there or in Kazakhstan). Bombing the pipeline might not kill anyone, but it could create an environmental disaster. Given the often barren, mountainous, and remote nature of the territory that the pipeline will traverse, it will be next to impossible to patrol it all constantly. In effect, though helping to fulfill China’s demand for petroleum, investment in Kazakhstan increases Chinese security concerns. A disruption of supply and shipments would obviously have very serious consequences for the Chinese economy; this, in turn, could seriously affect the regime’s legitimacy. The central government will have to ensure the security of the pipeline through all of Xinjiang and be deeply concerned about its security in Kazakhstan. The logical consequence of increased dependence on oil originating in Kazakhstan is tightened security in Xinjiang, as well as the perception that China’s security interests extend well beyond the Chinese border into Kazakhstan. This, in turn, may encourage a buildup of PLA forces in Xinjiang, especially along the border, in areas with large ethnic-minority populations.
Security concerns set a general context for PRC policy in Xinjiang. Both for its own petroleum (and other raw materials) and its links to external sources that will be increasingly vital for the Chinese economy, Xinjiang looms increasingly large in China’s geo-economic and geo-strategic thinking. This will require greater investment in and development of the infrastructure in Xinjiang and greater integration of Xinjiang into the Chinese heartland. Although Xinjiang is likely to be an economic beneficiary of Chinese energy policies, it will be more firmly supervised and controlled by the center, which will focus its concerns on the suppression of opposition and protest in the XUAR.
ECONOMIC DEVELOPMENT IN XINJIANG
As stated above, Xinjiang has done relatively well economically since 1949, and it is becoming increasingly essential in the eyes of the central elite to keep Xinjiang part of the PRC. Although energy plays a particularly important role in this today, the desire to integrate Xinjiang into the national economy and national transportation and communications networks stretches back to the early years of the PRC, as is suggested by comparative statistics.
Table 5.1 shows the national rankings for per capita capital construction (investment) and per capita income for the provincial-level units of the Northwest and for China’s provincial-level autonomous regions (Ningxia and Xinjiang fall into both categories). The data suggests that except for in the period 1971–1975 (the Fourth Five-Year Plan), Xinjiang has ranked among the top ten provincial-level units in the country in terms of per capita investment. Qinghai has done even better, and Ningxia also has done well. Guangxi, the most populous autonomous region, has fared poorly, even in the reform period after 1978, despite its nominal “coastal status.”11 Shaanxi, Gansu, and Inner Mongolia were major recipients of investment during the Mao period, in part because they were important centers of defense industrialization. Little can be said about the data on Tibet, except that much of it is missing.
In terms of per capita income, the data are less enviable. Xinjiang ranks in the top half of all provincial-level units. However, in recent years, the rankings for per capita income of the northwestern provinces and all autonomous regions have declined, compared to other provincial-level units, though perhaps some limited Open Door effects are seen in Guangxi’s recent relative improvement.
Other things being equal, per capita investment should, perhaps with a lag effect, generate roughly comparable per capita income, as investment is a major determinant of income. Here, the differentials between investment and income rankings suggest a number of possibilities: first, investments were used very inefficiently; second, significant amounts of investment may have been devoted to improving the basic infrastructure and not to directly productive activities; and/or third, given the lower levels of economic development in the provincial-level units of the Northwest and in China’s provincial-level autonomous regions, the same amount of capital generates a lower rate of return than in the more developed regions. It is likely that all three hypotheses are true to some extent, but there is no obvious reason why Xinjiang seems to outperform Qinghai, to say nothing of the other provincial-level units considered here. Is there any reason to believe that more investment went to directly productive activities in Xinjiang than in the other provincial-level units? Or that Xinjiang can use investment more efficiently than other autonomous regions or units in the Northwest? Or that Xinjiang invests less in infrastructure? The answer to all these questions is probably no. However, though Xinjiang’s superior performance vis-à-vis other northwestern provincial-level units (and other autonomous regions) cannot be fully explained, its poorer performance in terms of the effect of investment on income compared to “core” Chinese provinces can be explained by the higher levels of preexisting infrastructure in those provinces, their smaller geographical size, their more plentiful rainfall and therefore more productive agriculture, and probably their more highly educated populations.
TABLE 5.1
Average National Rank: Capital Construction per Capita and GNP per Capita
Sources: Quanguo ge sheng, zizhiqu, zhixiashi lishi tongji ziliao; Zhongguo guonei shengchan zongzhi hesuan lishi ziliao; Zhongguo guding zichan touzi tongji ziliao, 1950–1985; Zhongguo guding zichan touzi tongji nianjian, 1995; Zhongguo tongji nianjian (1996, 1997, 1998, 1999).
Note: This table uses rank orderings because I have not been able to locate a consistent series of data for investment and gross domestic product using the same prices.
*=77–80
It might be objected that per capita rankings skew the results because, with the exception of Guangxi, the provincial-level units under consideration are among the least populous in the PRC. Even Guangxi is not among the ten most populous provincial-level units. There is some validity to this point, but aggregate investment levels or rankings based on all provinces as equal units would be more distorted. Is it reasonable to compare the fewer than 5 million people in Ningxia with the 90 million or so in Henan (which would be one consequence of comparing aggregate provincial investment or incomes)? Tibet, Ningxia, and Qinghai, the three least-populous provincial-level units, are highly unlikely to have aggregate investment or income rankings above the lowest three. Per capita figures are more illuminating regarding investment priorities and income results because if aggregate figures were used, the gross disparity between large and small provinces would obscure shifting emphases in the allocation of capital and in provincial incomes.
Assuming that these figures are valid, we are left with the question of why Xinjiang has done better in terms of investment and per capita income than the other autonomous regions or provinces in the Northwest. As the rankings indicate, this pattern of results is not simply the product of the reform program (though market allocation processes may explain why other provinces have clearly seen their rankings drop off).
Unfortunately, no good answer suggests itself, and Xinjiang’s superior rankings are not easily explained. We might speculate that the independence of Soviet Central Asia has convinced Beijing of the need to build up Xinjiang, or there might be something unusual about the nature of economic development in Xinjiang, such as the important role of the PCC. These ad hoc hypotheses remain problematic. There really isn’t much difference between the investment rankings for Xinjiang during and after the Mao period. At the same time, there is nothing obvious in the nature of the PCC that makes it a more efficient user of investment and generator of revenue than other enterprises or organizations in China.
We are left with the fact that Xinjiang, in terms of the national rankings for per capita investment and per capita income for the entire period from 1953 to 1998, had the best results among the northwestern provinces and autonomous regions. There is no explanation for this. Although the overall pattern of results lends some credence to the view that the coastal provinces are getting rich at the expense of the interior provinces, at least in terms of investment, the evidence is not conclusive.
Map 5.1. Percentage of Xinjiang Population Classified as Han by County, 1998
But though Xinjiang’s performance is satisfactory, provincial-level rankings hide huge disparities in the distribution of gross domestic product at the county and city level in Xinjiang. Map 5.1 shows the distribution by county-level unit of Han in Xinjiang in 1998. The areas where Han are in the majority are in the north and east (or generally north and east of the Tianshan mountain range). In other northern areas where Han are not the majority, such as in the Ili district, they are nevertheless a substantial minority. The only area where Han are the majority south of the Tianshan mountain range is in Aqsu municipality. In many counties in the south and west, especially in the Kashgar and Hotan districts, Han make up less than 10 percent of the population.12 Map 5.2 shows the distribution of the Uygur population in Xinjiang; clearly Uygurs overwhelmingly predominate in the south and west.
There has been a general stability in the composition of county populations, at least according to the population data in the Xinjiang Statistical Yearbook. Comparing the population data for 1988 and 1998 reveals that only one area switched from a bare Uygur minority to a place where no ethnic group was in the majority. However, that was Yining, an area of great unrest. The number of Uygurs as a percentage of the overall population in Xinjiang declined slightly, while the percentage of Han increased minimally. There was a minor increase in the percentage of Han in the Uygur heartland, in southwest Xinjiang, particularly in the Kashgar district. Nonetheless, in most cases, Han remained less than 20 percent of the population. Such figures, however, do not include migrants whose housing registration (hukou) was not transferred with them.13
Map 5.2. Percentage of Xinjiang Population Classified as Uygur by County, 1998
Map 5.3 shows that the areas where per capita annual income is below ¥4,000, or less than two-thirds of the regional average, are concentrated in the southwest. In three regions (prefectures)—Kizilsu (Kezilesu), Kashgar, and Hotan—there are no Han-majority counties or cities and no areas with per capita income above ¥4,000. This area comprises twenty-four cities and counties, or more than a quarter of the eighty-seven cities and counties of Xinjiang. In 1998, forty-five of the eighty-seven major sub-region jurisdictions had per capita incomes below ¥4,000. Han were the majority in only four of these areas, meaning that forty-one of these areas were ones where ethnic-minority peoples constituted the majority of the population. In the entire region, thirty-two counties and cities were Han majority. As noted, only four of those units had per capita incomes below ¥4,000. Forty-one of fifty-five counties and cities where ethnic-minority people were the majority of the population (or almost 75 percent of such areas) had per capita incomes below ¥4,000. In only two counties where ethnic-minority peoples were the majority of the local population was the county per capita income above the regional average. Those two were Urumqi County, which surrounds the regional capital (45 percent Han), and Shanshan County, the location of a major coal mine and an oil field (31 percent Han). Fourteen of the Han-majority counties and cities had per capita incomes above the regional average of ¥6,229. Thus, there is a very strong correlation between areas of Han majority and high per capita income (and conversely, areas with a large percentage of non-Han and low per capita income). Moreover, it is likely that this is not just a correlative relationship but a causal one.
Map 5.3. Xinjiang Per Capita GDP by County, 1998
Han economic dominance was manifest in other areas as well. In the 1995 PRC industrial census, 191 large- and medium-sized enterprises in Xinjiang were identified (some key enterprises related to national security may not have been reported). Of those 191, 180 were state-owned industrial enterprises. In none of the large enterprises (40 in number) were their managers unambiguously members of ethnic minorities (one was surnamed Ma, which is often a Hui surname but not always). Of the 151 medium-sized enterprises, the managers of 2 were unambiguously minority, 3 were surnamed Ma, and 1 was unclear (someone named Xuanyuan Guoxin).The rest were almost certainly Han.14 Only 6 of the 40 large enterprises were in ethnic minority–dominant areas (all in and around Yining). Two were PCC enterprises, and all 6 were apparently managed by Han. With the exception of one enterprise in Korla and another in Ruoqiang County (also a PCC enterprise) no large enterprises were located south of the Tianshan mountain range. Eighteen of the 40 were in and around Urumqi, and no other place had more than 3 large enterprises. Medium-sized enterprises were somewhat more broadly distributed, but they remained concentrated in the north as well. Fifteen of the 151 medium-sized enterprises were located in ethnic minority–dominant units. Only 6 were south of the Tianshan mountain range: 3 in the Korla district, 1 in Kashgar, and 2 in Hotan. Four enterprises were in Han-dominant Aqsu (2 of them were PCC enterprises). Fifty-nine of the 151 medium-sized enterprises were located in Urumqi. Overall, large- and medium-sized enterprises accounted for almost three-fifths of industrial output in Xinjiang in 1998 (a year of major decline in Xinjiang industry. In 1997, large- and medium-sized enterprises accounted for two-thirds of industrial output).15
Since the overwhelming majority of large- and medium-sized enterprises in Xinjiang are state owned, their location, in part, reflects conscious state choice. In addition, as state-owned enterprises, managers are appointed from the appropriate nomenklatura lists. The lists for centrally “owned” enterprises are unlikely to contain the names of many ethnic-minority members. Those for PCC enterprises, of which there are 49, are also unlikely to contain many names that are not Han, given that the membership of the PCC is about 90 percent Han. Thus, it is not surprising that for the most part, non-Han are not factory managers. Indeed, the distribution of large- and medium-sized enterprises in Xinjiang and the selection of managers who are overwhelmingly Han reflect deliberate state choices. For whatever reasons, the state has not invested heavily in western and southern Xinjiang, and these areas remain predominantly agricultural, poor, and non-Han. Consequently, the privileged position of Han in Xinjiang’s political economy is strongly reinforced by decisions concerning allocation.
Although China has been engaged in economic reform since the late 1970s, Xinjiang looks more and more like a centrally planned region with a “traditional” colonial economy. In 1997, almost 68 percent of all industrial output came from large- and medium-sized enterprises.16 Also in 1997, 1,288 of 7,634 industrial enterprises in Xinjiang were controlled by the center. At a time when central control over industrial enterprises has dropped dramatically, such a large number of centrally controlled enterprises is extraordinary.17 Whereas nationally, in 1997, the state sector produced about a third of all industrial output in the country, in Xinjiang, it produced more than half of the state’s industrial output. Overall, the state sector accounted for three-quarters of all industry in Xinjiang.
Except in 1959, 1975, and 1976, heavy industry never constituted more than 55 percent of the total output of light and heavy industry combined during the Maoist period. Only in 1959 did heavy-industrial output exceed 60 percent of total industrial output. However, in 1995, heavy-industrial output was 64 percent of total industrial output; in 1996, it was 67 percent; in 1997, it was almost 69 percent; and in 1998, it exceeded 69 percent.18 In short, in recent years, Xinjiang’s industry has been oriented more toward heavy industry than ever in its past; in other words, it has become more economically imbalanced. Such a concentration on heavy industry is usually equated with a planned economy, but for China as a whole, even in the heyday of state planning, heavy-industrial output did not consistently rise above 60 percent of the total output of light and heavy industry combined.
The reason for this increasing concentration on heavy industry is that natural-resource development, especially petroleum products, has been the core element of central-government plans for Xinjiang for some time, and for statistical purposes, natural-resource development is by definition a branch of heavy industry. Heavy industry is more capital intensive than light industry, is often less labor intensive, and has higher embodied technology, which means that it requires a more educated workforce. In most cases, natural-resource development, especially on a large scale, is a task for the central government or a central corporation, as opposed to being a task for a regional or provincial government or below. This is particularly true of petroleum. In statistical terms, extractive and raw-materials industries accounted for almost 60 percent of industrial output in Xinjiang in 1997.19
However, with the independence of Soviet Central Asia, there was some hope that Xinjiang might profit from the Open Door Policy in the way that coastal provinces had. (The coastal provinces had switched to labor-intensive light-industrial production and encouraged extensive exports, often by foreign-funded enterprises.) Foreign trade has grown substantially throughout the entire 1978–97 period in Xinjiang, at a rate of about 25 percent per year, which exceeds the growth of GDP (about 11 percent) by a significant margin. However, it appears that in the 1990s, the rate of growth of total foreign trade has slowed, and that imports have been growing substantially faster than exports. As table 5.2 shows, total foreign-trade growth has slowed substantially, especially since 1995, and the growth of exports has been slower than the growth of regional GDP.20 Imports have grown dramatically, especially in the early and mid-1990s. In 1996 and 1997, Xinjiang had a negative trade balance. In the five years (1994–98) for which statistical information is available, the border trade has been in deficit as well. Although a number of the border traders doing import business may be PRC citizens, it appears that trade and export promotion have not played a major role in Xinjiang’s economic development. Moreover, almost 75 percent of all of Xinjiang’s foreign trade came from Urumqi and the PCC.21 It would be hard to argue that foreign trade will be a major stimulus to economic growth in Xinjiang, especially in the poorer areas where ethnic-minority populations predominate.
In recent years, the key agricultural priority in Xinjiang has been to promote the cultivation of cotton, and Xinjiang is the country’s leading cotton producer.22 Cotton is one of the few agricultural crops in which China’s price structures approximate international market conditions. If international prices are set equal to 100, for China as a whole, prices for rice average 101, and cotton, 107. For most other crops, China’s prices are considerably higher. With WTO accession, much of Chinese agriculture faces severe competition.23 China has vast amounts of cotton in warehouses, and supply grossly outstrips demand. Farmers in Xinjiang are not making a profit on cotton.
Of equal importance to the economic difficulty of raising cotton in Xinjiang are the potential long-term ecological effects. A direct parallel can be drawn between efforts to promote cotton production in Xinjiang and the “virgin lands” program promoted by Nikita Khrushchev in the 1950s and 1960s. The program pushed for extensive land reclamation in Soviet Central Asia, with the new fields to be used primarily to grow cotton. Water from the Aral Sea was to be the source of irrigation. The topographical and hydrological conditions were roughly similar to those in Xinjiang, though the Soviets may have had more available water than the PRC has in Xinjiang. However, the result in the former Soviet Union has been a massive environmental disaster. The Aral Sea is disappearing at an alarming rate. What was once the world’s fourth largest freshwater lake has lost between 50 and 60 percent of its volume; its salinity has increased from 1 percent to 11 percent.24 Since we lack precise hydrological information on aquifers in Xinjiang, we cannot make a definitive assessment on whether the water resources there are being used in a less exploitative fashion than was the case in Soviet Central Asia. It would be surprising, however, if long-term cotton production did not lead to similar kinds of environmental problems in Xinjiang.
TABLE 5.2
Xinjiang Foreign Trade (in US$1,000,000)
Note: For reasons not explained in the statistical yearbook, ordinary trade and border trade do not equal total trade or total import
Source: 1999 Xinjiang Statistical Yearbook (figures round up to the nearest million).
The major parts of Xinjiang where cotton is planted in large areas are in non-Han regions, especially the Aqsu and Kashgar districts. (Aqsu municipality has a Han-majority population, but the counties in the district do not. Since PCC farmers are 32.6 percent of the population in Aqsu municipality, without them, even Aqsu municipality would not have a Han majority.) Combining local (difang) production with PCC subregional production reaffirms this: together, Aqsu and Kashgar districts account for more than half of all the cotton acreage in Xinjiang. Although the locations of PCC subunit headquarters are known, it is unclear whether PCC subunit boundaries follow those of the political administrative divisions of the XUAR. It is therefore impossible to apportion precisely to the districts of the XUAR the acreage sown in cotton in PCC subunits.25
Thus, cotton is planted predominantly in areas inhabited by ethnic-minority peoples. It appears to be unprofitable and potentially environmentally unsustainable. It also appears to be promoted by the PCC in both northern and southern Xinjiang, and the amount of cotton acreage seems to be growing in tandem with the increasing Han population in northern Xinjiang.
Another indicator of ethnic inequality and the predominance of central-government authority in Xinjiang is in the budget. In the 1980s and 1990s, Xinjiang was a major recipient of central-government budget subsidies, with Beijing providing the equivalent of approximately 50 percent of expenditure in the region. The statistical materials do not explicitly demarcate the central government’s subsidies for earlier years, but 1966 was the last year that local revenues exceeded expenditures in Xinjiang. In theory, local governments are not supposed to run budget deficits, so the difference between local revenues and local expenditures (when there is a deficit) is a de facto form of central-government subsidy to the locality. Between 1973 and 1986, expenditures were at least 3 times local revenues in Xinjiang; from 1987 to 1997, they were closer to about 2.5 times local revenues. Expenditures in Xinjiang were thus highly subsidized, but since 1987, the rate of subsidization has been declining. Either the center has grown weary of providing such extensive subsidies, or the expenditures in Xinjiang are beginning to pay off with higher rates of revenue generation.26
It is also possible to look at where revenue was generated and where it was spent within the XUAR in 1997. In per capita terms, ¥323 of revenue was obtained through taxation and fees, and ¥528 was spent. (It should be noted that almost ¥4.5 billion of overall expenditure is unaccounted for, whereas only ¥200 million of overall revenue is not. Perhaps the PCC is not included in expenditures, but PCC budget figures are lacking.) In places where national minority peoples constitute a majority of the population, especially Ili, Aqsu, Kashgar, and Hotan, per capita expenditures were below ¥400, or about three-quarters of the provincial average. There were some Han areas with low per capita expenditures, for example, Shihezi and Kuytun, but both have high per capita GDPs. One area with a high percentage of minorities had a higher than average rate of per capita expenditure: Kizilsu, the least populated prefectural-level unit in Xinjiang. Han areas generally had higher per capita spending than the provincial average. It should be noted that in all minority areas, income was less than expenditure, indicating a certain degree of subsidy. Urumqi, Karamay, and Kuytun were the only places in the XUAR where income exceeded expenditures.27
The consequences of this pattern of expenditure are easily understood. Expenditures are used for many purposes, which include schools, public health, and economic development. If per capita expenditures in Ili, Aqsu, Kashgar, and Hotan were at best three-quarters of the regional average, it follows that in the most profound sense, life chances in those areas were substantially poorer than elsewhere in Xinjiang. The educational and economic disparities between Han and non-Han were made clear in a comparative analysis of the findings of the 1982 and 1990 censuses in China.28 There is no reason to assume that the situation has improved, and it may even have deteriorated further.
The meaning of these economic factors in Xinjiang’s developmental history and current status should be obvious. The central and regional governments appear to be pursuing a classic policy of economic imperialism, or internal colonialism, in the XUAR. The region is deeply dependent on the center for capital. This capital is used primarily to invest in the excavation and exploitation of raw materials. The center’s role in industrial ownership is also extensive. Investment is concentrated in heavy-industrial, raw-material sectors. Economic opportunities seem overwhelmingly to benefit Han, and there is a high correlation between above-average regional income and Han-majority populations in counties or cities. In general, there are higher rates of spending in Han-majority areas. The Han population tends to be urban; the ethnic-minority populations, rural. The government has apparently forced many farmers, a majority of whom are from national minorities, to pursue a cash crop, cotton, rather than grain or other forms of agriculture that may be better suited to local conditions.
Many characteristics of the reform period in the coastal areas are not present in Xinjiang. Despite the independence of the Central Asian states, Xinjiang has not seen a rapid growth of regional exports in recent years. It does not have a large private or collective sector. It does not have a balanced pattern between light and heavy industry. It is not basically reliant on itself for its investment and expenditures.
This is not to say that other forms of colonialism are not present in Xinjiang as well.29 However, economic colonialism may be as important as they are, and it may serve as the basis upon which other forms are based. Economic colonialism may also have consequences for political control and political action.
In a recent essay, Barry Sautman challenged the view that Xinjiang should be seen as an internal colony.30 Sautman’s argument is too complex to be addressed fully here; however, to an important degree, it hinges on intention—the intention of the colonizers to exploit the colonized. He argues that many outcomes in Xinjiang can and should be explained by general Communist Party policies toward the countryside, by geography and climate, and by “the politics of anti-separatism,” which keeps the Han in charge. But this last point undermines his argument. He notes that all the minority areas of China are part of China because they were conquered by the Han core. But in my opinion, it is the Han presence alone that keeps Xinjiang, Tibet, and other regions in the PRC, and PRC leaders have pursued policies designed to tie border and minority regions to the Chinese heartland. These policies have sometimes been relatively benign. In aggregate economic terms, Xinjiang has done reasonably well. It does not appear, however, that the central government has gone out of its way to support minority areas in Xinjiang, and both deliberately and unconsciously, its policies have reinforced a pattern of dominance by which the Han disproportionately benefit. To my mind, such a pattern of rule is characteristic of imperialism and internal colonialism.
The Han project to turn Xinjiang into an internal colony has been at least partially successful. Over the last twenty years or so, northern Xinjiang (Dzungaria) has been largely integrated into the PRC. Much of northern Xinjiang is composed of Han-majority administrative divisions. There has been substantial economic development in this area. But though the Han presence has undoubtedly been greatly magnified in northern Xinjiang since 1949, it is far from clear that the national minority groups are reconciled with political and economic integration. As it was throughout much of the 1900s, the Ili region appears to be less than happy at the prospect of deepening integration into China. The Han are not a majority in the Ili district, but they are present in substantial numbers, and one should not doubt the extensive presence of the People’s Liberation Army and the People’s Armed Police in Ili. Nonetheless, Ili has remained a center of protest and violence throughout much of the history of the PRC (and of prior Chinese regimes as well), and it has been perhaps at the core of what the central and regional authorities call “splittist activities” in the 1990s. Economic imperialism and internal colonialism may have achieved the center’s aims in Xinjiang, but they have not reconciled relations between Han and non-Han. Quite the contrary, they may have solidified ethnic identifications that resist either economic or political accommodations and perpetuate cycles of protest, violence, and repression.
THE XINJIANG PRODUCTION AND CONSTRUCTION CORPS AND MIGRATION
The Xinjiang Production and Construction Corps is a state within the state in Xinjiang. It is widely seen as comparable to Tang and Qing dynasty institutions that created farms in border areas manned by soldiers. These encampments would increase the Han presence in these regions, provide manpower with some military training to defend these sensitive areas, and above all, extend China’s imperial domain to areas that were strategically significant but sparsely populated, thus providing a buffer between the nomadic tribes of Central and Inner Asia and the Chinese heartland. When PLA troops entered Xinjiang in 1949, their commander was Wang Zhen. During the Yan’an period, Wang commanded the 359th Brigade in Nanniwan, south of Yan’an. Mao had extolled the unit for its combination of military skill and political consciousness and for its willingness to engage in extensive agricultural land reclamation.31 Inspired by a similar strategic logic imbued with socialist consciousness (as well as by more immediate concerns, such as the difficulty of returning PLA and surrendered KMT troops to the Chinese heartland, and perhaps also by Wang Zhen’s careerist interests), the PCC was informally created in 1952 and formally established in 1954.32 It was and is an overwhelmingly Han organization.
During the 1990s, the PCC constituted more than an eighth of the entire population of Xinjiang, and its membership (about 2.4 million) was about 90 percent Han. Since 1990, it has had the equivalent of provincial status in terms of planning (jihua danlie). It controls a very large proportion of the productive assets of Xinjiang, including about 20 percent of arable land devoted to grain production and more than a third of land under cotton. In 1995, at the time of the Third National Industrial Census, the PCC had 2,859 industrial enterprises and production units. This constituted an increase of 639 units since the industrial census of 1985. The output of these enterprises was more than 22 percent of the total industrial output of Xinjiang in 1995.33 The PCC produced more than a third of all the sugar, cloth, canned goods, and machine-made paper in Xinjiang. In 1985, the ratio of heavy industry to light industry was about 7 to 3. In 1995, it was 82.5 to 17.5. In 1995, the PCC controlled 48 of Xinjiang’s 195 large- and medium-sized enterprises. This 25 percent figure was a significant decrease from the 1985 census, when 32 of the region’s 89 large- and medium-sized enterprises were controlled by the PCC.34
The PCC’s economic centrality in Xinjiang has been obscured in recent years by attacks on its “labor education” and “reform through labor” camps. Although the number of these penal institutions is not known, the PCC has an unusual range of power over judicial issues, which demonstrates that though it is an economic, military, and administrative institution, it has attributes of territorial administration that are unlike those of any other institution in China.35
The PCC serves as a vanguard of Han penetration into Xinjiang. As the figures in table 5.3 suggest, the PCC constitutes a significant proportion of the Han population in southern and western Xinjiang. It is recruiting new members, and it has been encouraging poor Han from interior provinces to migrate to Xinjiang to work the cotton fields. It has military functions and serves as a backstop to PLA and armed police forces in the region.
Although the PCC has been encouraging migration into Xinjiang, it is also true that the economic opportunities for migrants in Xinjiang have themselves stimulated migration. Certainly it appears that the Xinjiang authorities have done little or nothing to discourage Han from migrating into the region. If migratory flows are ranked in four tiers (from most to least in terms of the number of immigrants) Xinjiang ranked in the second tier (with only the coastal provinces in the first tier).36 Legal (hukou-transfer) migrations averaged between 75,000 and 100,000 per year in the 1990s.37 Figures for floating-population (illegal) migration are at best estimates and are highly inexact. Several sources suggest figures for total in-bound migration in the range of 250,000 to 300,000 per year. All are assumed to be Han migrants.38
TABLE 5.3
Xinjiang Administrative Subdivisions: Population, GDP, and Ethnic
*FPCC is the agricultural population of Production and Construction Corps units in the area. Agricultural members of the PCC constituted about 75 percent of the PCC’s total membership.
Source: 1999 Xinjiang Statistical Yearbook
In official statistics, Han make up a little less than 40 percent of the population in the region. But this probably significantly understates the number of Han in Xinjiang. It probably excludes PLA and armed police-force members who are not originally from Xinjiang. There are no readily available figures on PLA and armed police deployments in Xinjiang, but given Xinjiang’s size, sensitivity and importance, and security problems, these forces are likely to be in the 250,000 to 500,000 range, again with the vast majority being Han. In addition, if the flow of migrants is in the order of several hundred thousand per year, and has been for several years, it is not unreasonable to believe that Han may constitute a majority of the population living in Xinjiang at any particular moment.
CONCLUSION: POLITICAL EFFECTS AND THE LEGACY OF HAN COLONIALISM
One of the most often remarked features in theories of international imperialism is the link between local colonial elites and elites in the metropolitan power. For a variety of reasons that do not need to be explored here, nontrivial numbers of ethnic minorities have been co-opted to serve the PRC administration of Xinjiang. For Han leaders and Han elites, they are, in many if not all cases, token examples of what the Hans would like to project as the positive state of ethnic relations in China. For minorities who want little or nothing to do with China, ethnic elites in the military, the government, and the CCP apparatus are traitors.
These local ethnic leaders must conform to the vision of official nationality relations, and they dutifully repeat the official line to outsiders. Thus, in a meeting with foreign reporters, the regional governor (a Uygur) was quoted as saying, “It is good for Nanjiang [southern Xinjiang] for Han to be sent there. They meet the needs of development.”39 The mayors of Yining and Urumqi (both Uygurs) relayed similar views to foreign reporters on a number of occasions. (That foreign reporters were allowed an extensive visit to Xinjiang in the early fall of 2000 suggests the leadership felt the security situation was well under control. Nonetheless, stories filed by these reporters make clear that Uygur antipathy toward Han efforts is not hard to find, despite the often quite controlled nature of these press visits.) The official line implicitly accepts the view that the Uygurs in particular are too poor and ignorant to carry out economic development themselves. Only with an influx of Han migrants (or implicitly, total acceptance of Han/PRC ways by the national minorities) can Xinjiang rise. The fact that the center is investing heavily in Xinjiang is proof of Han/PRC good intentions and impartiality. The remedy for splittism is explicitly stated to be economic development. But this goes hand in glove with extensive suppression of the national minorities, the closing of mosques and Islamic schools, and other actions that make it harder for Uygurs and others to organize for collective action, even if such actions reinforce non-Chinese identities.
Probing the true thinking of minority elites in Xinjiang is impossible. Nonetheless, they are likely to play a key intermediary role between the central-government elites and local Han leaders. Behind closed doors, they may be able to adjust policy marginally and make Han more conscious of the consequences of their policies. Incremental changes of this nature are surely not likely to win support from those who are unreconciled to Han/PRC control. To exercise such possible influence, minority elites must echo the official view in public. They may be perceived as loyal tools of Han imperialism. Yet they may doing what they can to make that imperialism more tolerable, perhaps because they recognize that Han control is not going to go away.
The size of this stratum of minority officials in Xinjiang is hard to determine. Their presence, as long as they remain players in the colonial game, provides evidence for the myth of the “Chinese civilizing project.” Ultimately, their example may be powerful enough to attract more minorities into accepting Han/PRC rule.
What can be said about the prospects for independence, or “splittism,” in Xinjiang? There is no unified opposition and no widely agreed upon leader who is seen internationally (and even in China) as speaking for Uygurs or Xinjiang in the way that the Dalai Lama speaks for Tibet. Moreover, it is not clear that other minority peoples in Xinjiang would welcome a Uygur-based state. The proponents of independence have been willing to use violence to pursue their ends in ways that have greatly raised the costs of Chinese rule, and they have been met in return with fierce repression.40 The disorienting effects of economic growth are undermining the “traditional” ways in much of Xinjiang, and divisions among and within minority groups are growing. As one traveler recounted,
Young Uygurs, educated in Chinese institutions, were following the Han in accepting Western pop culture and mores. Islam forms the basis of Uygur identity, to be sure, but modernizing, secular layers are gradually accruing. Xinjiang’s Uygurs may well come to feel as torn between East and West as many Turks are today. Rent by contradictions between Islamic traditions and acquired Western mores, and split along educational and generational lines, they will be more easily controllable by Beijing.41
Others appear to dispute this view. But what cannot be disputed is that Xinjiang (along with Tibet) is a core component of China in Han eyes. Increasingly, it is seen as vital for the economic future of China. The Chinese government has invested much more in Xinjiang than it has in Tibet, both in per capita terms and in the aggregate (as table 5.1 shows), and the per capita income in Xinjiang ranks considerably higher. Thus, the PRC will not allow an independent Turkistan to become an option. This is likely to harden non-Han opposition and make Han rule in the short to medium term more difficult and costly, in terms of human rights as well as finances. But over time, the influx of Han, the co-optation of local people, and rising incomes will limit the effectiveness of any calls for independence. For the most part, Uygurs and others will not be assimilated, but increasingly, Han China is giving them a choice: participate in the process of Chinese rule and do better economically or resist and be suppressed (and be economically ignored). This is not an attractive choice. It is hard to think of a time when other options were available to either Han or non-Han in Xinjiang. But short of a political collapse in the Chinese heartland causing the Han hold on Xinjiang to slip, China has demonstrated the will and the power (and the diplomatic skill) to impose its control over Xinjiang. It will not be easily displaced.
POST–SEPTEMBER 11 DEVELOPMENTS
The Government of the PRC quickly signified its support for the U.S. War on Terrorism in the aftermath of the terrorist attacks in the United States on September 11, 2001. This cooperative approach was linked with efforts by the Chinese government to brand Uygur organizations in general as terrorist and to associate them with the Taliban and Osama bin Laden in particular. At the rhetorical level, in mid-January 2002, the State Council Information Office issued a lengthy document laying out a litany of alleged terrorist activities undertaken by Uygur independence organizations. Although many specific acts of violence were detailed, no supporting evidence was provided. “Incomplete statistics” cited in the report reveal that 162 people were killed in Xinjiang between 1990 and 2001, and more than 440 were injured. Interestingly, Uygurs and other non-Han who worked at low levels in the state administration seem to have been a special target for attack and assassination, with more than a dozen examples provided (or close to 10 percent of the incomplete statistics.)42 Yet no attacks occurring after September 11 were reported in the document, and despite the purported ties of separatist forces to the Taliban and other external sources, knives and explosives, rather than guns and other weapons, seem to have been used most often in “terrorist” activities.
Publication of the State Council document had consequences: Muslim clerics were required to demonstrate their allegiance (biaotai) to the Chinese state, and religious activity was strictly controlled and monitored by the ever-present police and military forces. China continued its efforts to extradite Xinjiang opposition figures from the Central Asian states (the State Council document reported that a dozen or so had already been returned to China); Beijing received a promise from Afghan interim leader Hamid Karzai that Afghanistan would return any Muslim separatist with Chinese citizenship as part of China’s $150 million aid package to post-Taliban Afghanistan.43 All the while, repression, extensive well before September 11, continued unabated.
These actions serve to illustrate the Chinese view that Xinjiang is an inalienable part of China and that the state and its agents will do whatever it takes to incorporate Xinjiang into China more fully. Opportunist actions, such as linking the Uygur opposition with September 11, serve to contain U.S. and human-rights organizations’ criticisms of China’s ongoing suppression of minority groups. At the same time, its efforts to expand economic development in Xinjiang continue, with Han in the vanguard. Chinese aid to Afghanistan comes with significant strings attached. Every possible means is used to further China’s control over its largest “autonomous” region.
NOTES
I owe a special debt to Neil R. Taylor for making the maps. I gratefully acknowledge the assistance of Morris Rossabi and the other conference participants.
1. Goncharov, Lewis, and Xue, Uncertain Partners, 121.
2. Holloway, Stalin, 177; Li et al., Dangdai Zhongguo de hegongye, 19–20; Fu, Dangdai Zhongguo de Xinjiang, 393–94, 451.
3. MacMillen, Communist Power, 23–24, 122, notes the presence of Soviet consulates in Urumqi and the Ili district.
4. See Fu, Dangdai Zhongguo de Xinjiang, and Zhongguo Gongchandang Xinjiang Weiwuer Zizhiqu for general background.
5. The case of Tibet is very well known. On Inner Mongolia, see Woody, Cultural Revolution in Inner Mongolia. Guangxi’s experience during the Cultural Revolution is also known to have been very violent, including reported cases of cannibalism.
6. Zhongguo binggong nianjian, 594–98. Information on the location of these enterprises comes from the 1995 Industrial Census. The Third Front was a Maoist policy of military industrialization that began in 1964. Its goal was to build weapons plants in the Chinese interior so that when one or both superpowers invaded and occupied the industrialized heartland, the Chinese people could continue to resist by fighting a “people’s war.” For details, see Naughton, “Third Front.”
7. For more detail on Guangxi, see Guangxi guofang gongye.
8. McMillen, “Xinjiang and Wang Enmao,” esp. 574–81.
9. For PCC membership figures, see Xinjiang shengchan Jianshe Bingtuan 1998 tongji nianjian, 39. Data on heavy and light industry and central subsidies comes from Quanguo ge sheng, zizhiqu, zhixiashi lishi tongji ziliao and various Xinjiang statistical yearbooks.
10. On petroleum development in Xinjiang and some of its implications for internal and external security, see Christoffersen, “China’s Intentions,” and Krekel, “Cross Border Trade.”
11. Despite its having a coast, most discussions of coastal provinces exclude Guangxi, and Guangxi has been less centrally involved with (and received fewer benefits from) the Open Door Policy, a policy that signaled China’s renewed political and commercial relations with the rest of the world after 1998.
12. My spelling of subregional political units in Xinjiang generally follows the rendering found in Zhongguo diminglu. Where these are not obvious, I also transliterate the Chinese characters in pinyin. The exceptions are Uygur (Uighur), Kashgar (Kaxgar), Aqsu (Aksu), and Hoten (Hotan).
13. Xinjiang tongji nianjian (1989), 60–65; Xinjiang tongji nianjian (1999), 60–65.
14. Zhongguo Disanci Gongye Pucha ziliao guangpan.
15. Xinjiang tongji nianjian (1998), 347; Xinjiang tongji nianjian (1999), 347, 480.
16. Xinjiang tongji nianjian (1998), 347, 468.
17. Ibid., 347.
18. Quanguo ge sheng, zizhiqu, zhixiashi lishi tongji ziliao, 923.
19. Xinjiang tongji nianjian (1998), 347.
20. Xinjiang tongji nianjian (1999), 347, 480.
21. Xinjiang tongji nianjian (1998), 571, table.
22. See Becquelin, “Xinjiang in the Nineties.”
23. Yang, “‘Ru shi’ hou.”
24. See Stone, “Coming to Grips”; Micklin, “Desiccation of the Aral Sea”; and Abdiiaev, “Disaster Zone.”
25. See Xinjiang Shengchan Jianshe Bingtuan 1997 nianjian, inside cover.
26. Quanguo ge sheng, zizhiqu, zhixiashi lishi tongji ziliao, 944.
27. Xinjiang tongji nianjian (1998), 189–91.
28. Hannum and Xie, “Ethnic Stratification.”
29. See, for example, Gladney, “Representing the Nationality”; Dikotter, Discourse of Race; and Harrell, “Introduction.”
30. Sautman, “Is Xinjiang an Internal Colony?”
31. Selden, The Yenan Way, 251–54.
32. For background on the PCC, see McMillen, “Xinjiang and the Production,” 65–96.
33. Xinjiang Shengchan Jianshe Bingtuan 1997 nianjian, 5–8 and Zhonghua Renmin Gongheguo 1995 nian, 1.
34. Zhonghua Renmin Gongheguo 1985 nian Disanci Quanguo Gongye Pucha ziliao huibian, 572.
35. See Seymour and Anderson, New Ghosts, Old Ghosts, 44–71, and Xinjiang Shengchan Jianshe Bingtuan 1997 nianjian, 112–31.
36. Gu Chaolin, oral presentation at the China Colloquium, University of Washington China Studies Program, Seattle, Wash., October 2000.
37. Based on figures from Xinjiang tongji nianjian (1990–1999).
38. Tayler, “Foreign Affairs,” 4; Pomfret, “In China’s Wild West”; and “China: Xinjiang Receives Rising Numbers of Migrant Farmers.”
39. Kwang, “Outpost Set to Rise,” 4
40. Amnesty International, “Gross Violations of Human Rights.”
41. Taylor, “Foreign Affairs,” 3.
42. Information Office of the State Council, “‘East Turkistan’s Terrorist Forces.”
43. See Zhongguo Xinwenshe, “Xinjiang Mobilizes”; Xinhua News Agency, “Chinese News Agency Says”; Agence France-Presse, “Muslims Placed under Tight Control”; and Agence France-Presse, “Karzai Agrees to Repatriate.”