Testimony Before the U.S.-China Commission
December 7, 2001

“Calculating China’s Military Expenditure”*

Testimony Provided by: David Shambaugh
Professor of Political Science & International Affairs
Director, China Policy Program
Elliott School of international Affairs
The George Washington University & Nonresident Senior Fellow
Foreign Policy Studies Program, The Brookings Institution

Introductory Remarks

It is a privilege and honor to have the opportunity to testify before the U.S.-China Commission at its hearing on “Chinese Budget Issues and the Role of the PLA in the Economy,” and I welcome this opportunity to explore this complex subject and to try and answer the questions of Commissioners.

There are few areas of Chinese military affairs more opaque and difficult to research than the revenue/expenditure and budget/finance domains—but perhaps none more important to understand. PLA doctrine, force structure, threat perception, and organization are all areas which are more transparent and researchable than PLA economic affairs. Yet, questions associated with the PLA's fiscal base are crucial to understanding these other areas. The allocation of financial resources is indicative of strategic priorities and calculations. In the case of the PLA, it is important not only to know where the money is going, but also where it comes from—as the military has always had extra-budgetary sources of revenue (yusuanwai) and significant hidden categories of expenditure in other ministerial budgets or secret accounts. Clearly the official defense budget is only a fraction—approximately one-half in my best judgment—of the total revenue accruing to, and expended by, the PLA. It is therefore vital for analysts to think in terms of the total revenue base rather than simply allocated budget when assessing the PLA's finances, i.e. the total pool of funds available to the PLA.

The Commission has, in its letter of November 5, asked me to address a number of key and important questions concerning China’s military budget and defense expenditure. I wish I could answer them all with clarity and accuracy. I will do my best, and believe that my lengthy and extensive research in this area permits me to probe more deeply and more accurately than most—but unfortunately many of the questions posed are ultimately unanswerable, at least with the degree of precision and empiricism one would want and expect. The lack of official transparency in Chinese military expenditure remains a serious impediment to research and understanding—yet, as is seen below, much can be ascertained from sources published in China.

Despite the opaqueness of China’s military expenditure and the consequent difficulties of doing research on the subject, I wish to draw to the Commission’s attention to the fact that there exists a wide variety of “open-source” information published in China in Chinese that is of considerable research and intelligence value. Unfortunately, in my personal view, the U.S. intelligence community has thoroughly failed to tap into, and exploit, these open-source materials. If nothing else, this Commission can make a major contribution to U.S. national security and American intelligence if it strongly recommends that a concerted effort is made to collect, translate, and publish these open-source materials about the Chinese military. This recommendation applies in particular to the translation of books (or parts of books), as this is where the vast majority of valuable open-source material about the PLA is contained. Periodicals also contain valuable information (and, here, it seems that the intelligence community and Foreign Broadcast Information Service are doing a slightly better job).

With respect to Chinese military expenditure, there exists a surprisingly large number of materials published in China on defense economics (guofang jingji), military finance (jundui caiwu), military expenditure (jundui feiyong or junfei), defense expenditure (guofang feiyong or junfei), and logistics work (houqin gongzuo) on which to base research into this subject area. The field of defense economics has, in fact, emerged as a bona fide field of teaching, research, and study in China.¹ These sources contain numerous nuggets of information and a surprisingly systematic picture of the defense budget process and sources of revenue (although it is notably weak on precise figures). One must comb this literature widely—which is only published and available in China—in order to piece together a reasonably coherent picture of the PLA's multiple sources of revenue and areas of expenditure. No doubt, the PLA itself does not know the full extent of its earnings and expenditures, but if foreign analysts tap the primary sources insights can be gained. The “black box” of the budget process, the revenue base, and expenditure parameters can all be illuminated by a careful reading of these sources.² In addition, interviews with active and retired PLA personnel in China and abroad can supplement the documentary data.

The Impact of Reform

Like other facets of the PLA, the financial arena is another undergoing comprehensive reform. There have been four principal areas of military fiscal reform process in recent years.

The first concerns the commercialization and subsequent divestiture of commercial assets by the PLA. In the mid-1980s the military was authorized to go into commercial business (bing shang) as a means to offset and compensate for low levels of state allocations to the PLA. The process of commercialization worked as intended—in fact much better than intended. Various PLA units set up a plethora of enterprises and commercial activities. This helped top-up paltry army coffers, but also had the very deleterious effect of soldiers spending time in unprofessional business activity (much of it illegal) instead of training, diverting military resources for commercial purposes, and creating a significant “second economy” in China. Thus, after the negative effects of military commercialization became apparent in the late-1990s, the government and Central Military Commission (CMC) issued several orders banning PLA business activities. None succeeded, and the problem worsened. Only after the joint State Council, Chinese Communist Party, and CMC order of July 1998, did the PLA’s commercial involvement truly begin to be reduced.³ Following the 1998 order, the commercial divestiture process passed through several phases, which are detailed later in this chapter.

The second reform was the regularization (zhengguihua) of accounting and auditing procedures in the PLA, beginning in the early-1990s. This move also met with resistance, particularly as military units sought to hide their assets and profits derived from their commercial activities. But, gradually, a regular auditing system was introduced from top to bottom of the system.

The third principal reform has been the marketization and rationalization of the defense industrial procurement system. A major institutional reform was inaugurated in 1998 with the creation of the General Armaments Department (Zong Zhuangbei Bu), or GAD, and reorganization of the Commission on Science, Technology, and Industry for National Defense (COSTIND). The goal of this reform was to make the entire defense industrial and scientific/technological sectors more efficient and cost-effective. This had certain implications for the military budget process, as a more market-based procurement bidding system was introduced. Also, following the 1999 Yugoslav War, the military was given a substantial boost in its allocations for weapons and other equipment—coming on top of double-digit real post-inflation increases that the PLA had enjoyed since 1989.

The fourth area of reform was introduced in 2001, and concerned the very process of PLA budgeting. In a radical initiative Zero-Based Budgeting (ZBB) was introduced. Since 1997-98 ZBB was introduced to many ministries and entities under the State Council, on the order of Premier Zhu Rongji. This initiative was part of a package of phased fiscal and accounting reforms intended to make the government’s entire fiscal system more efficient and accountable. China’s socialist economy was filled with accounting and budgeting irregularities that allowed for double-counting, hidden assets, and a variety of off-budget revenue and expenditure. The PLA’s fiscal management was no different, and probably worse. Under ZBB, all units were required to draw up their anticipated expenditure for the next fiscal year from zero, rather than from the previous system of taking last year’s expenditure and adding to it by a certain percentage for next year. With this reform, “Units no longer are supposed to arrange their budgets on the basis of their base figures of the previous year…. They begin to calculate and examine their annual budgets and itemized budgets from ‘zero,’ with the arrangement of their budget [requests] in order of priority.”⁴ One important item apparently left out of ZBB initiative are personnel costs.⁵

The remainder of my written testimony is divided into four principal sections. The first section offers some summary comments on the official defense budget in recent years. The second section looks more carefully at the defense budget process and system. The third section attempts to account for areas of expenditure under the official defense budget and the PLA's off-budget revenue and expenditure. It attempts to estimate total PLA revenue and expenditure. The final section examines aggregate trends.

The Official Defense Budget

There is no doubt that official Chinese defense expenditure is rising, and rising at double-digit rates since 1989 and in real terms since 1997. Since 1989 China’s official military spending has risen from 21.8 billion RMB in 1988 to 141 billion RMB in 2001. The 2001 figure is approximately equivalent to $17 billion US dollars. Table 1 below reveals the official military budget during this period, the percentage of increase year-on-year, and the percentage of the military budget as a percentage of total Central Government Expenditure (CGE).

Table 1: Official Chinese Military Expenditure, 1988-2001

As a percentage of central government expenditure, China has shouldered a significant “defense budget burden” over time. While China expended a relatively high level of its central government budget on the military over time, an average of 16.7 percent from 1950 to 2001,⁶ this percentage “burden” has fallen to approximately half that amount (8.5 percent) during the period since 1988. As a percentage of GNP, however, China’s defense burden has been more modest. From 1950 to 1980, China’s official national defense expenditure constituted approximately 6.35 percent of national income on average, but it dropped dramatically to approximately 2.3 percent for the 1980s, and fell even further to an average of 1.4 percent during the 1990s.⁷

Ultimately, of course, it is a question of what the money is being spent on and what the relative allocations indicate about intentions. This is discussed at greater length below, but suffice it to note here that personnel costs have always absorbed a large percentage of Chinese defense expenditure. From 1950 to 1970 personnel expenditures accounted for 40 percent of official defense expenditure, dropping to 30 percent during the 1970s, and rising again to approximately 40 percent during the 1980s.⁸ During the 1990s approximately 35 percent of the defense budget was devoted to personnel costs (salaries, housing, medical care, dependent’s support, etc.).

It is also interesting to note the fluctuations in annual defense expenditure. In most cases China's defense budget reflected external tensions, but in others it reflected domestic considerations. Over the first forty years of the People's Republic we see a close correlation between China's external threat environment and defense spending. In a couple of cases (the Great Leap and Cultural Revolution) austere domestic conditions produced a shift in defense expenditure, but, on the whole, defense expenditure paralleled China's security environment and posture.

In the decade since 1989 a demonstrable break from the previous pattern has been witnessed. During a period when China’s national security environment has arguably never been better and there is no pressing external threat, its defense spending is rising sharply. Some in the United States point to these twin trends as indicating China’s hostile intent and pursuit of a crash military modernization program that threatens the interests of the United States and its allies in the Asia-Pacific region. I believe this view is not warranted by either the aggregate level of expenditure or the amount devoted to equipment allocations. The PLA is undertaking a complex and long-needed process of modernization—a process which is not tantamount to a buildup of military forces.

Accurately estimating PLA expenditures is a notoriously difficult process, fraught with frustration owing to the lack of transparency on China’s part.⁹ Although the PLA adamantly maintains that its announced official defense budget constitutes its total military expenditure, it is widely accepted by foreign analysts that the official defense budget figure constitutes only a fraction of the total revenue available to the People's Liberation Army and falls far short of actual expenditure. Western estimates range from two to twelve times the announced official budget. The data and analysis presented below concurs with this foreign skepticism, but my estimates fall at the lower end of this range—approximately one-half of total Chinese military expenditure. That is, I estimate that total PLA expenditure is 2-2_ times larger than the official budget. Because funds are often unspent and rolled over into the next fiscal expenditure cycle, funds for the PLA are buried in other budgets, and the military enjoys extra-budgetary commercial revenues, the total available amount of revenue available to the Chinese military is unknown—probably even to the PLA itself.

Categorizing Chinese Defense Expenditure

Any attempt to calculate Chinese military expenditure must begin with consideration of the categories of inclusion and exclusion in the Chinese defense budget. This is important because China does not apply the same categories that are standard in the international community—as used by NATO, the United Nations, the ASEAN Regional Forum, the World Bank, or leading international institutes such as the International Institute of Strategic Studies or Stockholm International Peace Research Institute. To be sure, these organizations are not themselves in complete agreement on appropriate categories—although they are all considerably more detailed than China’s official budget breakdown, and there is broad agreement on most categories.¹⁰ Some organizations, such as the International Monetary Fund and the U.S. State Department’s Arms Control and Disarmament Agency, employ a purchasing power parity (PPP) model, although their categorization is similar to the standard model employed by NATO’s.

Officially, China’s 1998 and 2000 Defense White Papers only distinguish three broad categories of expenditure: personnel, maintenance, and equipment. This elementary breakdown was only offered first in 1998, in response to considerable pressure from abroad for increased budget transparency; prior to this time only a single lump sum figure was provided. This source defines these three categories as follows:¹¹

• Personnel expenses: “mainly including pay, food and clothing of military and non-military personnel”;

• Maintenance expenses: “mainly including military training, construction and maintenance of facilities and running expenses”;

• Equipment expenses: “costs for equipment, including research and experimentation, procurement, maintenance, transportation, and storage.”

The 1998 White Paper goes on to say that, “In terms of the scope of logistic support, these expenditures cover not only active service personnel, but also militia and reserve requirements. In addition a large amount of spending is used to fund activities associated with social welfare, mainly pensions for retired officers, schools and kindergartens for children of military personnel, training personnel competent for both military and civilian services, supporting national economic construction, and participating in emergency rescue and disaster relief efforts.”¹²

China’s 1998 Defense White Paper claimed that, in 1997, 36 percent of official military expenditure was spent on personnel-related costs, while 33 percent was spent on operations and maintenance (O & M), and an additional 31 percent on equipment.¹³ The 2000 Defense White Paper revealed important adjustments in these relative allocations, as depicted below in Table 2.¹⁴

Table 2: Allocation of Chinese Military Expenditure 1998-2000

From these aggregate figures it is, of course, difficult to ascertain the percentage of expenditure on different branches and services of the PLA, but one authoritative source revealed that for the period 1950-1980, the PLA Air Force garnered an average of 31.37 percent of the defense budget, while the Navy got a meager 18.4 percent, leaving the ground forces to absorb the lion’s share of 50 percent of the budget.¹⁵ Of the spending on equipment for the ground forces, this source revealed the following breakdown: 17.22 percent for vehicles; 7.42 percent for tanks; 9.22 percent for communications equipment; 8.88 percent for ordnance and munitions; 7.5 percent for “other equipment,” 30 percent on personnel, and 20 percent on operations and maintenance.¹⁶ Of course, monies allocated for weapons development are buried in other budgets, but nonetheless these figures are illustrative of how the PLA spends its money. It is important to note that the PLA spends much more on people than arms.

Other Chinese sources go much further in defining and delimiting expenditure on and by the PLA. The most detailed and definitive is the Practical Encyclopedia of Chinese Military Finance,¹⁷ which lists fifteen separate categories and extensive subsidiary descriptions. Given the internal PLA classification of this source and the extensiveness of the volume (over 800 pages) this breakout of the official defense budget should be considered quite definitive. It also roughly corresponds to another key source: the 700-page Management of China’s Military Expenditure.¹⁸ From these definitive sources, intended for internal PLA usage, it is evident that the official PLA budget is quite comprehensive.

However, two key facts need to be borne in mind when evaluating these data: first, this constitutes only the centrally-apportioned percentage of defense expenditure and, second, several complete categories of funding fall entirely outside this budget framework. As noted above, the PLA has fiscally operated on a cost-sharing basis with sub-national governments, and units throughout the armed forces generate their own revenue from a variety of extra-military endeavors. It is unknown precisely how much of maintenance costs—particularly housing and food subsidies—or personnel overhead costs are met by non-central allocations, as these are not generally reported in provincial statistical yearbooks.¹⁹ Nor is it known, prior to the 1998 divestiture, how the estimated $600 million to $10 billion (low to high estimates) in commercial profits were distributed. But the important point here is that, while seemingly comprehensive in scope, the aforementioned categories cover only the central portion of military expenditures. However, they do illustrate the totality of expenditure categories covered by the official budget—thus revealing extra-budgetary categories and funds passed through other state budgets—and this in itself is an important discovery.

By international standards, the Chinese categorization includes some elements of expenditure not common, for example, to NATO, SIPRI, the IMF or World Bank.²⁰ Civil defense expenditure is normally excluded by these organizations, as are military pensions to veterans or demobilized servicemen (as they are considered transfer payments). Conversely, military aid, funds for paramilitary organizations, reserves and National Guard, all military-related R, D, T & E input costs, military space activities, and revenues from arms sales accruing to the military or its affiliated companies are all normally included in counting military expenditure—while none of these are included in Chinese categories.²¹ SIPRI specifically excludes some categories that have been regular components of the PRC military budget and PLA expenditure over time, such as veteran’s benefits and demobilization costs, funds for defense conversion and weapons destruction.

What Doesn’t the Chinese Defense Budget Cover?

From the above categorization it is clear that a fairly large range of defense expenditure costs are covered by the official central budget. These allocations are supplemented by allocations and revenue streams from sub-central governments, industries and factories, and commercial endeavors. By the late-1990s, individual units were thought to generate approximately half (in some cases more) of daily O & M costs through their commercial activities. This included, importantly, food production. The PLA also continues to manufacture a variety of its daily use equipment (which are, in fact, a form of “in-kind” payment). Salaries and a variety of daily maintenance costs are also topped up through proceeds from units' extracurricular activities. In addition, a variety of costs are paid for through other budgets or from off-budget revenue.

It appears that the official defense budget does not include all funds for: (1) indigenously made weapons and equipment production (as distinct from procurement); (2) some research, development, testing, and evaluation (RDT&E) costs; (3) the paramilitary People's Armed Police (Wu Jing), and reserves; (4) funds for special large weapons purchases from abroad; (5) funds directly allocated to military factories under the control of the GAD, and funds for defense industry conversion; and (6) military aid. How are these six categories of costs paid for? In addition, the PLA benefits from arms sales revenue and post-divestiture commercial revenue.

First it must be recognized that the PLA does not buy everything that defense factories produce. Some of these factories have converted to produce goods for civilian consumption. More to the point, however, in many of these factories production remains driven by socialist-style quotas or supply-side factors (e.g. maintenance of full employment) that are not responsive to consumer demand—including the PLA as principal consumer. For years the PLA has complained that it does not want to buy much produced by its own defense industrial system, but is forced to do so either for lack of alternative suppliers or because it is ordered to do so by the state. The defense industries share many of the burdens of other state-owned enterprises (SOEs).

When the ground, air, or naval forces seek to procure a given weapon system, these procurement (i.e. purchase) costs are apparently borne by the given service arm, as allocated through their annual appropriation as part of the defense budget. The revenue available for procurement, however, is fixed in the defense budget and calculated during the annual budget bidding process overseen jointly by the General Armaments Department and the Finance Bureau of the PLA General Logistics Department. When a service seeks to procure a given system, it contracts with the relevant ministry, which sub-contracts to the factories concerned. Before the 1998 reorganization of the defense industries (when competitive bidding was introduced) and creation of the General Armaments Department, the price paid for the hardware was fixed by COSTIND at an arbitrarily low level. Once prices were set and contracts signed, payment was made, apparently for finished items upon delivery. Thus procurement prices did not meet production costs, which must be borne by the defense industries concerned. Therefore, under the pre-1998 system the defense industries were largely responsible for their production costs while sharing R & D costs with COSTIND—the deficit was compensated/offset through direct subsidies to defense industries. It is too early to say how this has changed under the new system, which is still being worked out. While difficult to estimate, redundant and subsidized production in the defense industrial sector may easily amount to $1 billion annually.

The second area of defense expenditure not fully covered in the official budget is research and development (R & D). Estimating the channels and amounts of funding for this sector is a real conundrum. They appear to be derived from four sources—the General Armaments Department, COSTIND, the Ministry of State Science and Technology, and the defense industries themselves—although the division of labor and investment between each is unclear. Of the three, COSTIND has clearly been the principal source of R & D funds, although this is apparently changing with the creation of the GAD.²² However, a certain amount of expenditure is also paid through the separate line-item defense industry budgets (see below), although presumably this pertains to upgrading production technology (applied research) rather than basic research on systems design and performance. The latter is undertaken in a sprawling number of (numbered) military research institutes and factories. In some cases these institutes are affiliated with the ministerial defense industries and in others they are independent entities.

Following the 1998 reforms, COSTIND's budget now derives entirely from a specific line-item allocation from the State Council. A third source of R & D funds are allocated through the Ministry of State Science and Technology budget (MOST), although the percentage is unclear. Taken together, it would not be surprising if these extra sources added $1 billion to the military R & D expenditure every year.

The third area of defense expenditure not included in China’s official military budget pertains to the paramilitary forces and reserves. It is surprising not to find a line item in the detailed categories above for reserve forces, now estimated at 1.3 million and growing.²³ The costs of supporting these forces must be borne entirely by contributions of provincial and local governments. In this regard, it is odd to find costs for maintaining the militia included in the official budget (although, to be certain, a considerable portion of this expense is met locally). However, the omission of the 1.3 million-strong People's Armed Police (Wu Jing) is striking.²⁴ The Practical Encyclopedia of Chinese Military Finance explicitly states that the PAP is primarily funded directly from the Ministry of Finance and also through the Ministry of Public Security budget, although some sources indicate that it is partially paid for out of Ministry of State Security funds.²⁵ The PAP was formed in 1982 by combining units from the Ministry of Public Security’s Internal Security Armed Police, and the PLA’s Border Defense Corps, Gold Protection Corps, and Fire Corps. The Wu Jing is comprised both from recruits and from soldiers demobilized from the PLA. It is the state's first line of defense against internal civil unrest. In 1989 the Wu Jing proved totally incapable of handling the Tiananmen demonstrations, but has subsequently been retrained and rearmed. Thereafter they coped better with unrest by farmers, urban workers, and minorities. This is a high priority for China's leaders, and funding has followed. A related high priority is building up Special Police (Te Jing), which are being trained for rapid deployment, counter-terrorism, and other contingencies.²⁶ It is unclear how they are paid for, although probably through allocations to the PAP and the PLA ground forces (both of which maintain special forces). Wang Shaoguang’s research, based on the Ministry of Finance’s Public Finance Yearbook, reveals an allocation of 12.8 billion yuan for 1998, with an additional contribution of 0.334 billion yuan from local sources. This is, of course, unbelievable as it would result in average expenditure of roughly 1,100 yuan for every PAP soldier per annum, to say nothing of operations, maintenance, and equipment costs. Annual expenditures for the PAP on the order of $2-3 billion would be a reasonable estimate.

An important fourth category of spending outside the official defense budget is specially-earmarked for foreign weapons purchases. During the 1990s China bought an estimated $6.75 billion worth of weapons and equipment from Russia, on average $750 million per year. The cash portions of these purchases were paid for from a separate category of funds earmarked for foreign procurement by the Central Military Commission. Some of these early purchases, such as the first batch of Su-27 fighters, were paid for with one-third foreign exchange ($400 million) and two-thirds barter in consumer durables and agricultural goods, but after 1993-94 Moscow began to demand total payment to be made in foreign currency. Sukhoi-27 fighters cost about $32 million each “off the shelf” (and approximately 50 percent more for the kit assembly of 200 at Shenyang),²⁷ the Su-30 fighters were sold for approximately $47 million each, the Sovremenny destroyers cost approximately $1 billion each, Kilo submarines $350 million each, and the S-300 surface-to-air missile systems were sold for $500 million in 1995.

A fifth area of extra-budgetary allocations that have benefited the PLA are central allocations made directly to defense industries—although, strictly speaking, these factories and companies are not owned by the PLA. These large subsidies are paid directly by the State Council to defense industries or those factories that produce partially for the military (e.g. electronics), but are administratively under one of the ten State Council corporations. Many weapons production costs are thus defrayed by the State Council through its subsidies to the relevant defense industry/corporation, rather than being carried in the defense budget. These corporate (ministerial) budgets are not made public in the Finance Minister's annual budget speech, nor are they available in the Tongji Nianjian (Statistical Yearbook). These direct line-item allocations to the defense industries could easily amount on average to $500 million per defense corporation (some more, some less) or $5 billion collectively for the ten major defense corporations that exist today—not to mention the extracurricular earnings by the corporations themselves.

An important subset of this category of extra-budgetary subsidies are funds allocated for defense industry conversion and earnings by these industries.²⁸ Nearly 70 percent of the output value of military factories is now accounted for by production of civilian goods, and during the Seventh Five-Year Plan (1991-95) the State Council earmarked 6 billion yuan ($1.14 billion) for facilitating conversion. That amount has declined during the latter half of the 1990s, and probably only amounts to approximately $500 million in the Ninth Five-Year Plan (2000-2005). To be sure, most of China's estimated 50,000 defense industrial factories—which employ up to 2.5 million employees—have not converted successfully. Hence they require substantial state subsidies. One report noted that 50 percent of defense industry production capacity remains idle, and described such factories as "an unbearable burden on the national economy." ²⁹ Wang Shaoguang estimates that these subsidies to loss-making military enterprises and for conversion amounted to four billion yuan in 1998.³⁰ Thus one can assume that conversion subsidies, while considerably down from the early-1990s, still amount to approximately $500 million per annum.

Finally, China still provides military aid to a handful of Asian and African states.³¹ Much of this goes to Pakistan, Burma, and Bangladesh and comes in the form of training of officers at the PLA National Defense University and military academies (“tuition” is usually paid for in entirety by China), and technical assistance accompanying arms transfers (see below).

While these seem to be the principal areas of military-related expenditure falling outside the scope of official defense budget categories, the importance of cost-sharing through other state budgets and local government allocations cannot be overemphasized. The PLA still apparently adheres to the “three-thirds” policy whereby many personnel and maintenance costs are proportionately split among central, provincial, and local governments.

Pensions and demobilization of servicemen and women are a prime example. From 1987 to 1997 the PLA demobilized approximately 1.2 million troops.³² In 2000 it concluded another round, demobilizing an additional 500,000 service personnel. It is estimated that approximately 10 percent of these demobilizations have been officers. The costs of demobilization have been substantial—particularly for the high-ranking officers who require/demand large pensions and perquisites. They are permitted to maintain their salaries, plus retirement bonuses and pension, housing, travel funds, free health and hospital care, and often a car and driver. Lower-ranking officers and enlisted personnel receive a one-time demobilization payment. These costs are covered partially in the official defense budget, but are also paid for through the Ministry of Civil Affairs budget, which is responsible for civilian cadre retirements as well. County, municipal, and local governments also underwrite, directly and indirectly, a large amount of the associated costs. Central Military Commission Chairman and Communist Party General Secretary Jiang Zemin, as well as senior military officials, have frequently commented on the need to give "high priority" to these demobilizations. Wang Shaoguang estimates, based on a survey of public finance and provincial statistical yearbooks, that annual non-central allocations to demobilized personnel amounted to 3.6 billion yuan in 1998,³³ but many of the subsidies are disguised. Since pensions and demobilization costs are not usually included as defense expenditures, I do not include demobilization and pension expenditures in the off-budget spending of the PLA.³⁴ The “three-thirds” policy also applies to housing and related garrisoning costs, as well as local contributions to energy expenditure (fuels) for some units. If one accepts that the figure provided in the 1998 Defense White Paper that 38 percent of official military expenditure is spent on personnel costs, and that this amount constitutes the central government’s allocation, under the “three-thirds” policy provincial and local governments would then spend $8.99 billion (at official exchange rates) in “matching funds” in the 2000 fiscal cycle. This amount should then be added as a category of extra-budgetary expenditure.

Two other sources of revenue deserve mention: proceeds from commercial activities and arms sales.

In the 1990s, PLA Inc. (as it became known abroad), cashed in on its comparative commercial advantages.³⁵ These came not only from converted defense industries, but also from a wide range of investment and production schemes. The PLA owned some of China's prime real estate, and has leased it out at high rents. Many local airlines were owned and managed by PLA front companies. Numerous hotels and guest houses were opened. The PLA Construction Corps charged localities hefty fees for heavy construction work previously undertaken for free as a symbol of the Army's selfless devotion to the people. The military capitalized on mines under their control by selling metals and minerals at market and above-market prices. The PLA's once-elite hospitals began to admit those who could afford the price of admission.³⁶ Virtually every military unit set up one form or another of cottage industry, and many were involved in joint ventures with foreign entities. PLA ships, planes and other modes of transport were put to commercial use (and became involved in smuggling rings). “PLA Inc.,” reached its prime in the mid-1990s, when somewhere between 15,000-20,000 companies were known to exist. Their assets ballooned and profits were plentiful. One Chinese source claims that total PLA assets totaled RMB 180 billion ($20.2 billion).³⁷ Foreign estimates of annual profit ranged from $1-3 billion, although the General Logistics Department claimed in 1998 that it was on the order of $600-700 million per annum.³⁸ While the 1998 order to the PLA to divest its commercial investments has proceeded remarkably well, it has certainly not been complete. It is estimated that as many as 20 percent of the units involved in extracurricular commerce have carried on their activities—which would yield approximately $2 billion per year in revenue.

The vast majority of the earnings of PLA enterprises remained with the unit that generated them, and did not make their way into the General Logistics Department's budget stream. They helped to defray local operating costs and compensate for the inadequate allocations from the Center. The commercialization of the PLA saved many military units from destitution at the very time that their central allocations were being drastically cut back, but the PLA's rapidly growing involvement in the market economy was having a deleterious effect on military professionalism.³⁹ Morale was low and recruitment difficult as soldiers earn considerably less than merchants or peasants.⁴⁰ Concomitantly, the ranks are filled with soldiers anxious for demobilization in order to take advantage of business opportunities. Commercial priorities ran at cross-purposes with the corporate ethos that the PLA High Command was trying to instill. Incidents of corruption, speculation, profiteering, smuggling, illicit sales and purchases, and other "unhealthy tendencies" became widespread.

The final category of PLA revenue is sales of weapons abroad. Proceeds from arms sales are not normally a category counted on international defense budget ledgers, as such funds normally accrue to private sector defense contractors who build and sell a given weapon system abroad, but in market economies and most countries defense industries are not government-controlled. Not so in China. Some of China’s principal arms export companies are attached to one or more General Departments of the PLA, while others are State Council entities.⁴¹ Normally, proceeds from arms sales are paid directly by the foreigner purchaser to the export company concerned—once production, storage, and transport costs are recovered, the company concerned is supposed to remit one-third of the profits made to its parent General Department, one-third to the CMC, while keeping one-third for itself.⁴² In practice, China’s arms export companies tend to keep whatever profits they can and probably obfuscate accounting books to conceal money made.

Arms sales provided China’s defense industries with an important source of extra revenue during the 1980s, but declined precipitously during the 1990s. At their height during the Iran-Iraq War, when China was selling to both sides, its arms export companies earned an average of $1.5 billion per annum in gross proceeds,⁴³ and garnered approximately $12 billion between 1985 and 1992.⁴⁴ According to ACDA figures, China’s exports topped out at $3.75 billion in 1988 but dropped to $0.58 billion by 1996.⁴⁵ China officially admits to selling $2 billion worth of weapons in 1987, acknowledges a drop to $900 million in 1991, and claims that the “volume of contracted business” did not exceed $1 billion in subsequent years through 1997.⁴⁶ Since 1998 revenue from arms sales has plummeted to approximately $600 million per year—with China’s major customers being the destitute nations of Pakistan, Bangladesh, and Burma. While these levels are not high, certainly when compared to other major arms exporting nations, they do represent an additional source of revenue flowing into the coffers of China’s Military-industrial complex. It is important to note that these earnings do not directly go to the PLA, although indirectly they benefit the PLA. Another way of stating this is that these monies benefit China’s armed forces and defense establishment, although not the PLA directly.


What Does It All Add Up To?

If these estimates of extra-budgetary sources of revenue are added to the 2000 official defense budget of $14.5 billion, one arrives at a total military revenue base of approximately $31.6 billion. These are depicted below in Table 3. This total would rank China third in the world in total military expenditure behind the United States and Russia, and just ahead of France, Japan, the United Kingdom, and Germany.⁴⁷

Table 3: Estimated PLA/Military Establishment Expenditures (2000)

There are surely some who will question and challenge these estimates as excessively high, while others will no doubt find them low. They are simply the most realistic—yet admittedly approximate—estimates I can offer based on knowledge of the extra categories of revenue available to the PLA and the likely amounts in each category. Some categories are fairly well known and accurate (foreign arms purchases, arms sales, military aid), while others are far less precise (subsidies, research and development, cost-sharing, commercial revenue). But the bottom line is that the PLA’s official budget presents only a part of the story.

As would be expected, the Chinese defense budget process has evolved over time. Perhaps it is more appropriate to say that it has devolved. Like much of the rest of the Chinese economy, central management and planning have been reduced since the 1980s, with responsibility for revenue generation falling increasingly with individual units at all tiers of the system. Like the civilian sector of the Chinese economy, the defense finance (guofang caiwu) system is in a half-way house between plan and market.

Chinese sources reveal three different processes for the assembling and promulgation of the defense budget: the centralized, decentralized, and combined systems (tongguan, fenguan, and jiguan).⁴⁸ It seems that during the 1990s all three systems operated simultaneously. The centralized system predominated, but was supplemented by the other two. Historically, the defense budgeting, allocation, and finance system has oscillated back and forth between a centralized Soviet-style distributive system and a more decentralized system. Key policies to decentralize funding sources and spending/auditing requirements came with the Great Leap Forward in 1958, the Cultural Revolution in 1966, and the Zhao Ziyang reforms in 1985, while efforts to centralize the system were made in 1952, 1954-55, 1965, 1978, and 1991.⁴⁹ The reforms of 2001 to institute Zero-Based Budgeting were intended, simultaneously, to decentralize the system again so as to increase unit accountability, while, also streamlining the method of allocation to give total discretion to higher-level authorities and preventing units from accruing off-budget sources of income. But the 2001 reforms do not change the structural htmects of the defense budget bidding process.

The centralized defense budget system is the one where central allocations are made to central, military and district levels. The process adopts a “down-up-down” system (zishang erxia) whereby the central-level GLD first works in conjunction with the Central Military Commission and Ministry of Finance to establish total expenditure targets, and then initiates a bidding system from military region/district levels, after which final expenditure figures and the central defense budget are set. The centralized military budget system cycle apparently works on an April to April fiscal year, and in an interactive vertical process between central, regional, and district levels. In March of every year, at the National People’s Congress, the annual national and defense budgets are announced by the Minister of Finance. This figure (for defense) should be viewed as both the culmination and initiation of the centralized defense budget process. That is, the total figure announced is both the outcome of a year-long bidding and negotiating process, as well as being the catalyst for the next budget cycle. The aggregate figure announced is, in effect, the total allocated pool of funds released by the State Council that military units can bid for to receive central allocations.

After the NPC, between April and June of every year, the GLD financial bureaus (GLD/FB) down to the division (and now brigade) level assess their needs and put together budget submissions for the coming year. This pertains to all ground force, air force, and naval units, but not the four general Departments (GAD, GSD, GLD, GPD), the Second Artillery, the Academy of Military Sciences, or the National Defense University—all of which are under direct control of the Central Military Commission. These estimates are passed up to military districts (still in the GLD/FB xitong) in July, where they are assembled and forwarded to the respective military region in August-September. In the early autumn the GLD/FB in Beijing begins the process of coalescing and assessing budget requests. When this process is completed, an annual "All Army Logistics Conference" is convened in Beijing, usually in November but sometimes December. Following the conference, final accounts are prepared by the GLD/FB for the past year's expenditures as well as the coming year's bids. These are forwarded to the Central Military Commission for consideration at year's end. At this stage seven central-level PLA institutions are required to submit their bids directly to the CMC (bypassing the GLD system): the General Staff Department; the General Political Department; the General Logistics Department itself; the General Armaments Department; the Ministry of Defense; the Commission on Science, Technology, the Second Artillery (nuclear forces); and military staff colleges directly under the control of the CMC (the AMS and NDU). Prior to placing COSTIND solely under control of the State Council in 1998, it too entered the budget process at this stage.

Thus, there is really a two-tier budget system for forces in the field and this group of central departments. To be sure, not all of these late budget bids by key central departments are built into the annual announced official budget figure. For example, as is seen below, a large amount of research and development costs are buried in other state budgets and/or are borne by individual factories in the defense industrial system. The Central Military Commission also has a large discretionary fund set aside specifically for purchase of foreign weapons systems, while People’s Armed Police, reserve, and militia costs are passed through other central and local budgets.

Usually after the annual November/December GLD conference, the Central Military Commission then collates total budget bids and determines what to allocate to each. These recommendations are then forwarded to the Ministry of Finance (Caizheng Bu) under the State Council, no later than the Chinese New Year (typically early February). The Ministry of Finance—in consultation with State Councilors, the Premier's Office, and certain Leadership Small Groups under the Party Politburo—then prepares final accounts for the previous year and comes to a final determination for military expenditure in the coming year. These figures are subsequently announced by the Minister of Finance in his annual speech to the National People's Congress in March. These figures form the basis of the centralized system, which then begins all over again.

In contrast to the centralized management system, the decentralized system (fengguan zhidu) is one where personnel and operating costs for military units are shared between general department, military region, and district levels. This system is informally referred to as the “three-thirds” system (san fenzhi san zhidu). In this system, military units at all three levels receive allocations outside the defense budget (yusuanwai) from provincial, municipal, county, and local governments. These contributions are themselves a kind of subsidy made by governments below the national level. For central-level units in Beijing and those at the military region level, provincial governments contribute funds, while the centrally-administered municipalities of Shanghai, Tianjin, and Chongqing contribute to locally-garrisoned forces. Beneath these, county, city, and township governments do the same at the military district and individual unit levels.

In reality, the “three-thirds” defense finance system (the combined centralized/decentralized one) has characterized the PLA budget process since the late-1980s. It is not yet clear, however, what the effect of the zero-based-budgeting system introduced in 2001 will be on the “three-thirds” system. Since the overall intention of the central government since the 1998 divestiture order has been to create a military budget system solely reliant on central government allocations, based on a rational budget submission process, it is quite likely that the contributions of non-central governments to locally-garrisoned PLA units will cease.

In all three systems, the General Logistics Department has traditionally been the key player in the defense budget process, particularly the GLD’s Finance Bureau (Houqinbu Caiwu Chu). While the GLD/FB plays a central role in the budget process, it is a coordinating role without any real decision-making power. It assembles the budget in consultation with the General Staff Department, service arms, and military regions, and forwards it to the CMC—which, in turn, negotiates the final annual budget package with the Ministry of Finance and State Council.⁵⁰ After the final budget is negotiated and set, the extensive nationwide GLD banking system (euphemistically known as the “logistics service bureaus” or houqin fuwu ju) allocates funds.

Needless to say, this entire defense budget process is fraught with intense lobbying. This usually takes place behind the scenes, but at the 1994 National People's Congress an unusual display of public lobbying took place. The PLA delegation introduced ten motions to the Congress aimed at increasing the defense budget. One of them proposed linking annual military outlays to growth in the economy and inflation by indexing. They argued for indexed increases of at least the inflation rate, and a year-on-year increase of 3.5 percent of Gross National Product (GNP).⁵¹ The PLA deputies who signed the petition (104 out of the 260-strong PLA delegation), pointed out that Chinese defense spending was only 1.7 percent of GNP—well below the needs of the PLA.

The idea of fixing the rate of annual increase in the defense budget, or indexing it to GNP growth or inflation, is an idea that began to surface in PLA circles in 1992-93 and continued throughout the decade. Frustrated that defense spending was barely holding even with inflation or actually declining in nominal terms, senior generals sought some way to insure adequate (i.e. indexed) annual increases. It seems that the idea of indexing arose out of frustration in the PLA High Command over a range of issues. There was much grumbling among officers that their pay had not risen to keep pace with inflation, and that, firstly, they deserved higher salaries. Because wages in the armed forces were so low, recruitment into the rank and file was becoming ever more difficult. Housing and other costs of billeting troops were woefully inadequate. There were reports of inadequate fuel and spare parts for training exercises. And new weapons systems were not receiving adequate investment.

Aggregate Trends

These estimates reveal revenue available to, and total military spending of, a little more than twice the official budget for the 2000 fiscal year. As noted above, this would place China third globally in aggregate defense spending, and very comparable to other medium-size powers like France, Britain, Japan, and Germany. But, while these other nation’s military budgets have been declining in the post-Cold War period (the “peace dividend”), China’s military spending has been steadily rising. While China’s military expenditure was closely correlated to its external threat environment for the first forty years of the PRC, during the last ten years, when there has arguably been no pressing external threat, Beijing’s military spending has soared. It has doubled in real terms in a decade. The official budget has risen at double-digit rates for twelve consecutive years since 1989—with an average annual increase of 15.5 percent. To be sure, during 1993-1997 China suffered inflation at approximately the same level (thus nullifying the increases), but for the periods 1989-92 and 1997-2000 China’s economy suffered deflation—thus, for much of this period, the increases have been real and substantial. While China continues to spend a high proportion of net government funds available in the annual budget (an average of 17 percent over the last fifty years and 8.6 percent from 1989-2000) relative to other countries, its “defense burden” remain modest in terms of the percentage of GDP spent on defense (approximately 1.4 percent of the official budget). Even if this is tripled to allow for the extra-budgetary revenue available to the PLA, this would only put China in the league of the United States (in percentage terms) and far below the Cold War levels of the former Soviet Union (which spent nearly 20 percent of GDP on defense). While increasing at a rather substantial rate over the post-1989 period, it is also not accurate to identify any kind of “crash” buildup of the military. China is simply not spending excessive amounts of available funds on the military, nor is there any evidence of heavy investment into particular programs (except perhaps short-range ballistic missiles since 1996 and cruise missiles since 1998).

The extra monies for the military are also going much further due to the significant reduction in personnel. This has permitted an increase in personnel expenditure and improvements in salaries, housing, and troop maintenance—thus permitting the military to recruit and retain better-educated soldiers and officers. Most of these reductions have come out of the ground forces—thus permitting an increase in recruitment in to the Air Force and PLA Navy. Mothballing of antiquated equipment in recent years (particularly aircraft and armored vehicles) has also saved considerable money, spare parts, and personnel. In other words, as the PLA has downsized it has become more rationalized and cost-efficient. In addition to freeing up funds for personnel, it has also increased monies available for procurement at home and abroad. The purchases of advanced equipment from Russia and Israel are indicative of this new liquidity, but it will become particularly apparent when new indigenous systems begin to come on stream around 2004-2007 (fighters, surface combatants, and submarines). But if the PLA still spends nearly 40 percent of its budget on personnel, this does not leave a great deal to invest in procurement and R & D. Of course, this is where extra-budgetary funding comes in. We noted above the variety of funds generated by units and firms that are ploughed back into the defense industrial system. These amounts have been reduced as the PLA has proceeded to largely divest itself of its commercial business empire.

Finally, it should be noted that the revenue from China’s arms sales has plummeted over the past 15 years. Simply put, Chinese weapons are a last resort for most developing nations, and Beijing’s failure to compete at all in the international arms market is testimony to the pathetic state of China defense industries (see Chapter 6). China’s military aid also plummeted over the same period of time.

In sum, the PLA has more money available than ever before, is spending it in a much more rational manner, but there still is scant financial evidence of a significant military build-up that constitutes a “China threat.”

FOOTNOTES

*This written testimony is drawn largely from my forthcoming book Modernizing China’s Military: Progress, Problems, and Prospects. All rights pertaining to the content of this testimony is reserved solely by the author and University of California Press.

1. The PLA runs a Military Economics Research Center in Wuhan (Junshi Jingji Yanjiuyuan), and there is a national Chinese Society for Defense Economics (with branches in various cities), and a number of periodicals devoted to the subject, e.g. Jundui Caiwu [Military Finance]; Junshi Jingji Yanjiu [Defense Economics Research]; and sometimes Zhongguo Jungong Bao [China Defense Industry News].

2. See, for example, People's University Reprint Series, Junshi [Military Affairs]; Lin Yichang and Wu Xizhi, Guofang jingjixue jichu [Basic Defense Economics] (Beijing: Academy of Military Sciences Press, 1991); PLA Logistics College Technology Research Section (ed.), Junshi houqin cidian [Dictionary on Military Logistics] (Beijing: PLA Press, 1991); Chinese Military Encyclopedia Editing Group (eds.), Jundui houqin fence [Section on Military Logistics] (Beijing: Academy of Military Sciences Press, 1985); Zhang Zhenlong (ed.), Junshi jingjixue [Military Economics] (Shenyang: Liaoning People's Press, 1988); Jin Songde et al, Guofang jingji lun [National Defense Economic Theory] (Beijing: PLA Press, 1987); Jiang Baoqi (ed.), Zhongguo guofang jingji fazhan zhanlue yanjiu [Research on the Strategy of China's Military Industrial Development] (Beijing: National Defense University Press, 1990); Gao Dianzhi, Zhongguo guofang jingji guanli yanjiu [Research on the Management of China's National Defense Economy] (Beijing: Academy of Military Sciences Press, 1991); Sun Zhenyuan, Zhongguo guofang jingji jianshi [The Construction of China's National Defense Economy] (Beijing: Academy of Military Sciences Press, 1991); Qiao Guanglie (ed.), Zhongguo renmin jiefangjun houqin jianshi [History of PLA Logistics Building] (Beijing: National Defense University Press, 1989); Wang Dangying et al, Guofang fazhan zhanlue yanjiu [Research on National Defense Strategy] (Beijing: National Defense University Press, 1988).

3. The order was titled “Removal of the Military, People’s Armed Police, and Political-Legal Units From Commercial Activities.”

4. For an explanation of this system in the PLA see Xiong Tingbin and Zhang Dongbo, “Central Military Commission Relays Plan for Reforming Compilation of Budgets for Armed Forces,” Xinhua Domestic Service, March 22, 2001, in FBIS-CHI, March 22, 2001.

5. My thanks to Christine Wong for making this point.

6. Author's calculation.

7. Contemporary China Series Editing Group (ed.), Dangdai Zhongguo Houqin Gongzuo [Military Logistical Work in Contemporary China] (Beijing: Zhongguo shehui kexueyuan chubanshe, 1990), pp. 306-307, and author’s calculations.

8. National Defense University Development Institute (ed.), Zhongguo guofang jingji fazhan zhanlue yanjiu [Research on China’s National Defense Economy Development Strategy] (Beijing: NDU Press, 1990), p. 243.

9. For an analysis of the problems associated with calculating Chinese defense expenditure, see “China’s Defense Expenditure,” in IISS, The Military Balance 1995/96 (Oxford: Oxford University Press, 1996), pp. 270-75. Also see Bates Gill, “Chinese Defense Procurement Spending: Determining Intentions and Capabilities,” in James Lilley and David Shambaugh (eds.), China’s Military Faces the Future (Armonk, NY: M.E. Sharpe, 1999), pp. 195-227.

10. For an excellent summary of the similarities and differences used by these different organizations see Somnath Sen, “Military Expenditure Data for Developing Countries: Methods and Measurement,” in Geoffrey Lamb and Valeriana Kallab, Military Expenditure and Economic Development: A Symposium of Research Issues (Washington, D.C.: World Bank Discussion Papers No. 185, 1992), pp. 1-18.

11. Information Office of the State Council, China’s National Defense, (Beijing: The State Council, 1998)..

12. Ibid, p. 82-83.

13. Ibid, pp. 84-85.

14. Information Office of the State Council, China’s National Defense in 2000, Xinhua, October 16, 2000.

15. Dangdai Houqin Gongzuo, op cit, p. 307.

16. Ibid.

17. China Military Finance Encyclopedia Editing Group, Zhongguo junshi caiwu daquan [Encyclopedia on China’s Military Finance] (Beijing: Jiefangjun chubanshe, 1993), part 3 (pp. 221-403).

18. Lu Zhuhao (ed.), Zhongguo junshi jingfei guanli (Beijing: PLA Publishers, 1995), esp. pp. 351-550.

19. If taken literally, the “three-thirds” policy would suggest a tripling of the official central defense budget figure, but Chinese sources insist that that this phrase indicates more a division of labor than division of actual running costs.

20. See Somnath Sen, “Military Expenditure Data for Developing Countries,” op cit.

21. See “Sources and Methods for calculating Military Expenditure,” SIPRI Yearbook 1999 (Oxford: Oxford University Press, 1999), p. 328.

22. See Harlan Jencks, “COSTIND Is Dead! Long Live COSTIND!,” in James Mulvenon and Richard H. Yang (eds.), The People’s Liberation Army in the Information Age, op cit, pp. 59-75.

23. IISS, The Military Balance 1998/99 (Oxford: Oxford University Press, 1999), p. 178.

24. Ibid,p. 181.

25. Zhongguo junshi caiwu shiyong daquan, op cit, p. 424.

26. For further discussion of Special Forces see chapter 4. These forces are known as Te Jing, but are not to be confused with the Special Police, which are SWAT teams subordinate to the Ministry of Public Security.

27. I am indebted to Ken Allen for this information.

28. See Paul Humes Folta, From Swords to Plowshares? Defense Industry Reform in the PRC (Boulder: Westview, 1992); Mel Gurtov, "Swords into Market Shares: China's Conversion of Military Industry to Civilian Production," The China Quarterly (June 1993), pp. 213-41; and Arthur Ding, “China’s Defense Finance,” op cit.

29. Jiang Baoji et al, "Lun wo guo guofang jingji tizhi mianlin de wenti ji gaige shexiang" [A Discussion of Problems Facing Our Nation's National Defense Economic System and Considerations for Reform], Junshi jingji yanjiu [Research on Military Economics], No. 12 (December 1990), p. 14.

30. Wang Shaoguang, “The Military Expenditure of China, 1989-98,” SIPRI Yearbook 1999 (Oxford: Oxford University Press, 1999).

31. According to SIPRI, military aid should be counted as an element of military expenditure. Since it is not formally listed as a component of the official defense budget, it is categorized here as extra-budgetary expenditure. See SIPRI Yearbook 1999, op cit, p. 328.

32. See Yitzhak Shichor, “Demobilization: The Dialectics of PLA Troop Reduction,” in Shambaugh and Yang (eds.), China’s Military in Transition, op cit, pp. 72-95.

33. Wang Shaoguang, “The Military Expenditure of China, 1989-98,” op cit.

34. See SIPRI Yearbook 1999 (Oxford: Oxford University Press, 1999), p. 328.

35. The best studies of the PLA in business are James Mulvenon, Soldiers of Fortune: The Rise and Fall of the Chinese Military-Business Complex, 1978-1998 (Armonk, NY: M.E. Sharpe, 2001); Solomon Karmel, China and the People’s Liberation Army (New York: St. Martin’s, 2000); and Tai Ming Cheung, Entrepreneurial Soldiers: The Chinese Army’s Quest for Profits, 1985-1999 (Oxford: Clarendon Press, 2001). Also see Tai Ming Cheung, “China’s Entrepreneurial Army: The Structure, Activities and Economic Returns of the Military Business Complex,” in C. Dennison Lane et al (eds.), Chinese Military Modernization (London and Washington, D.C.: Kegan Paul International and AEI Press, 1996), pp. 168-197.

36. Interview with military doctor in the PLA's famous 301 Hospital, April 1994.

37. Lu Zhuhao (ed.), Zhongguo Junshi Jingfei Guanli [China’s Military Budget Management] (Beijing: Jiefangjun chubanshe, 1995), p. 155.

38. Interview, GLD, December 8, 1998.

39. For an interesting exposition of the financial problems arising, and how the General Logistics Department is attempting to cope, see the symposium on military logistical work in Guofang Daxue Xuebao (February 1993), pp. 64-71.

40. See Ellis Joffe, “The PLA and the Chinese Economy: The Effect of Involvement,” Survival, Vol. 37, No. 2 (Summer 1993), pp. 24-43. .

41. Also see John W. Lewis et al, “Beijing’s Defense Establishment: Solving the Arms Export Enigma,” International Security (Spring 1991).

42. Interviews with NORINCO and General Logistics Department personnel, November 1993.

43. Arms Control & Disarmament Agency, World Military Expenditures and Arms Transfers, 1991-1992 (Washington, D.C.: ACDA, 1993), p. 100.

44. Richard Grimmett, Conventional Arms Transfers to the Third World, 1985-1992 (Washington, D.C.: Congressional Research Service, 1993), p. 60.

45. ACDA, World Military Expenditures and Arms Transfers, 1997 (Washington, D.C.: ACDA, 1998), p. 265.

46. China’s National Defense, op cit, p. 128.

47. International Institute of Strategic Studies, The Military Balance 1999/2000 (Oxford: Oxford University Press, 1999), pp. 300-302.

48. Wang Qincheng and Li Zuguo (eds.), Caiwu Daquan (Urumqi: Xinjiang renmin chubanshe, 1993), pp. 501-506. The following description of the three budget systems is drawn primarily from this source.

49. See the discussion in Long Youcai and Wang Zong (eds.), Jundui caiwu jianshe (Beijing: PLA Publishers, 1996), pp. 122-125.
50. This judgment is based on several published Chinese sources, interviews, as well as a CIA analysis—see James Harris et al, "Interpreting Trends in Chinese Defense Spending," in Joint Economic Committee of the Congress of the United States, China's Economic Dilemmas in the 1990s (Armonk, NY: M.E. Sharpe, 1992), pp. 676-84.
51. Jiefangjun Bao, 17 March 1994.