galbraith china hot money

8/14/2019 Galbraith China Hot Money

1/47

1

Is China Really Running a Trade Surplus?

James K. GalbraithProfessor

Lyndon B. Johnson School of Public AffairsThe University of Texas at AustinAustin, Texas

Sara HsuAdjunct Professor of Economics

St. Edwards UniversityAustin, Texas

Li JianjunProfessor of Economics

Central University of Finance and EconomicsBeijing, China

Last updated: December 30, 2007

The University of Texas Inequality Project

UTIP Working Paper 45

Abstract

We examine Chinas macroeconomic and trade accounts for simple, tell-tale signs that capitalinflows are being disguised as export earnings. We find large reported increases in a calculatedunit value of Chinese manufactured exports, which do not appear to correspond to increased unitprices in the accounts of countries importing from China. We therefore suspect that thelegalization of dollar accounts by firms resident in China, as well as an increase in expectation ofRMB appreciation which occurred in 2003, have led to large disguised capital inflows. Themagnitudes could range up to $529 billion by 2006. If this is correct, then China is not running a$170 billion current account surplus as officially reported in 2006, but rather a much smallersurplus, or even a deficit, obscured and financed by illicit inflows.


2/47


3/47

3

Capital Inflows into China

Hot money has been flowing into and out of China since its emergence as a major economy. Andthe Chinese government, which has a long history of capital control, is concerned about theissue. Since capital controls remain in force in China while the current account was liberalized in

1996

1

, it stands to reason that some efforts to evade capital control would flow through thecurrent account.

Officials at the Peoples Bank of China (PBOC) have confirmed that significant hot moneyinflows have run through the current account2 through the over-invoicing of exports, the practiceof overstating export values in order to bring foreign capital into the country. Hu Xiaolian,director of the State Administration of Foreign Exchange (SAFE), and vice governor of thePBOC, and Deng Xianhong, deputy head of SAFE, recently called for audits of short-termforeign exchange accounts to check these inflows (Xin 2007)3. An anonymous governor ofPBOC, associated with the National Development and Reform Commission (NDRC), hasestimated that "false exports during August-December 2006 resulted in an increase of US $17.5

billion in the favorable foreign trade balance, accounting for 17% of the total favorable balance(Zhong 2007).4 It appears that the Chinese government has estimated the occurrence and extentof capital inflows in the trade account by watching short-term foreign exchange transactions,upon which the PBOC imposed further regulations in February of 20075,

Scholars and bank economists have long watched Chinese hot money fluctuations, mainly via thecatch-all errors and omissions category in the balance of payments (Prasad and Wei 2005).Green (2006) estimates that hot money inflows comprised US $67 billion in 2005, although thisis a very rough estimate. Some have also suspected that hot money has also flowed into thecurrent account, taking the form of payments for fictitious exports or over-invoicing of actualexports. This phenomenon appears not to have attracted extensive scrutiny so far.6 Yet, it couldhave large implications for understanding the true trade and financial position of China today

In this paper, we investigate the extent to which capital inflows may have appeared deceptivelyin the trade account from 2003 to 2006. Not having access to the Chinese governmentsfinancial data, we take a simple alternative approach: we estimate the inflows using thepublished balance of payments data, while checking our estimates against investment, foreignimport prices, and financial activity.

1 See Li (2004). Capital Account Liberalization in China, The Chinese Economy, 37(1), pp. 85-116 for a timelineof current account liberalization.2 Wei and Zhang (2007) note that a government official of an anonymous country admits that capital inflows have

been introduced through the overbilling of exports. That the reference is to China is consistent with the context ofthis remark.3 Nineteen domestic banks and ten international banks have been punished for facilitating short-term money inflowsdisguised as trade or investment (Anderlini, Financial Times, June 27, 2007).4 On the other hand, Sun Mingchun, vice-president and Asia economist of Lehman Brothers Asia Ltd, stated that hotmoney inflows may be slowing down due to recently implemented checks on short-term capital inflows and stockmarket transactions (Zhang July 13, 2007).5 See Peoples Bank of China Adjusted Foreign Exchange Administration Policy towards Individuals6 Gunter (2003) makes a case for the phenomenon of capital flight from China, from 1984 to 2001, with over-invoicing of imports.


4/47

4

Chinas unlikely export values

A simple first cut at the problem involves making a large, problematic assumption: that changes in theinternal composition of Chinese exports within major product categories over a short three-year periodcan be safely ignored. (We will examine this assumption in detail later.) If we can treat the commoditycomposition within the major product categories as fixed, then dividing total export values by quantities

will give us a unit value measure of Chinese exports.

The official data available for this purpose are very erratic, but it is easy to show that this is due

mainly to the influence of two highly volatile export sectors, neither of which is quantitatively

very important7. We therefore construct a streamlined representation of export volume by

removing those sectors. For the remaining sectors, we find a steady increase in export volumes

over time. Dividing dollar volume by these quantities gives unit values8. We calculate unit

values using both OECD and CEIC data, applying the growth rate in calculated unit values after

2004 to quantities given in the OECD data9.

Table One shows export unit values from 1996 to 2006 calculated after excluding the volatile

sectors10. In manufacturing, these values are extremely stable through 2002, at around $0.39

USD per reported unit, and then they start increasing rapidly. In 2003, the manufacturing export

unit value jumps to $0.49, and in 2004, to $0.59. In 2005, the export unit value is $0.72, peaking

at $0.97 in 2006, and the overall export unit value follows the same trend, which is not surprising

since manufactures dominate Chinese exports.

Table One: Export Unit Value (In US Dollars)

1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006

Export UnitValue (Our

Calculations) 0.51 0.47 0.47 0.46 0.46 0.45 0.46 0.56 0.66 0.79 1.00Export Unit

Value (OfficialStatistics) 0.05 0.06 0.07 0.08 0.10 0.08 0.60 0.70 0.76 N/A N/A

ManufacturingExport UnitValue (Our

Calculations) 0.40 0.38 0.39 0.39 0.39 0.38 0.40 0.49 0.59 0.72 0.97

7 See Table Five below. The volatile sectors are Beverages and Tobacco (SITC 1) and the Special Commodities and

Transactions (SITC 9)8 This calculation is far from precise, because quantities are organized by type of unit, such as tons, thousands ofunits, and so forth, depending on sector, so there is no single consistent unit of exports. For this first cut, we are ineffect assuming the existence of a constant composite unit, not strongly affected by changes in the composition ofexports or by quality change.9 OECD data is the only data set that contains total quantities for SITC categories and for all trade, but only goes upto 2004. CEIC makes some quantities available, but not all. Therefore, we needed to use both.10 The World Banks WDI Online database includes information on Chinas export value indices. These are alsoshown to increase dramatically from 2003 to 2006. We choose not to use this information because data fromanother source, UN Comtrade, do not show the same increase.


5/47

5

Source: OECD and CEIC Data and authors calculations11

Now, let us suppose that Chinese exports had continued at unchanged unit values for the entireperiod from 1997 through 2006. What would have been the growth of total exports on that

assumption? Table Two gives actual export values, export values under the counterfactual of nounit value change, and the difference, which is attributable to changing unit values. It is clearthat a major part of the reported increase in Chinese exports is not due to increasing rawvolumes, but rather to some combination of reported price increases and product transformation,reported as rising unit value.

Table Two: Export Volumes and Total Exports Attributable to Volume Gain(in Billions of Current US Dollars)

Exports of

Goods andServices

Exports ofGoods andServices at

Unchanged UnitValues Difference12

2002 365.4 365.4 0

2003 485.0 398.4 86.6

2004 655.8 457.1 198.7

2005 836.9 490.2 346.7

2006 981.0 452.4 528.6

Source: World Bank WDI and authors calculations

As noted, part of the increase in unit values can be attributed to actual price increases in exports,and part to shifts in the composition of Chinese manufacturing output to higher technologygoods. But how much? That is the question we next examine.

Have Chinese manufactured exports increased in unitprice, that is, in value per item holdingcomposition and quality constant? If they had, we would expect that the unit import prices inmanufactured goods reported by other countries, especially Europe, Japan and the United States,would show comparable increases. However, as Tables 16 to 18 in the appendix demonstrate,

11 Rather than using given total quantities, the total of the individual commodity categories (minus the extremelyvolatile beverages and tobacco category) was used, since the two were not equal. Using given total quantities wouldresult in an even more dramatic increase in value per exported unit. The OECD does not yet have data for 2005-6.We therefore estimated this using quantities and values calculated from CEIC data, adding up quantities and valuesfor all categories that had quantities, and finding the unit value. We then looked at the growth rate of the CEIC unitvalues and applied this to CEIC data. The growth rate and unit value for 2004 was consistent with OECD data,providing an overlap in data sets.12 We calculated hot money inflows assuming they started in 2003. Therefore, we find that the difference in 2002is zero.


6/47

6

there is no sign of this at all. Every table shows stable or declining unit prices, and in some casessharply declining unit prices, for manufactured imports. Eurostat even publishes unit priceindexes specific to imports from China; these show no net change for manufacturing ormachinery, the major Chinese export sectors.

Further, if Chinese exporters hadsimply raised the prices of goods sold to the world market, wewould expect to see a loss of market share. Manufactured goods together total nearly 90 percent

of Chinese exports by value, and it seems unlikely that their prices could rise sharply without

significant losses in favor of Vietnam, Malaysia and other low-wage competitors. In fact, there

were no significant losses toward competing exporters. To the contrary, the quantity measures in

manufacturing show robust growth in 2003 through 2006. In the case of commodity-based

exports, such as animal and vegetable oils and fats, price increases could have happened without

loss of market share, insofar as overall commodity prices rose during this period. But such

goods are a trivial share of total Chinese exports by both value and quantity.

Table Three shows the ASEAN share of world markets over this period: there is no sign of an

increase that could be attributed to Chinese exporters pricing themselves out of the market.

Table Three: ASEAN Share in World Exports13

(Percentages)

1998 1999 2000 2001 2002 2003 2004 2005

ASEAN Share

in World

Exports 4.8% 5.1% 5.6% 5.2% 5.2% 4.0% 5.4% 5.4%

Source: UNCTAD Statistics Database

A second possibility is that China has upgraded the actual quality of its exports, justifying higherunit values not with price increases but with better goods. This possibility is particularly relevantto the processing trade, where China could be importing increasingly high-value goods in orderto finish them and export them again. But if this were the case, then unit values of Chineseimports in manufacturing would also be increasing. Table Four gives unit values for imports inmanufactures and machinery; no unusual increase is observed, although there appears to be a

steady slight progression in unit price from 2001 onward. This may account for part of theexport unit price increase, but not too much of it. This share of the unit value increases could bedue to an increase in the technological content of process-trade goods, which did occur startingin 2002. However, the fact that the unit value increase is not reflected in Chinese import unitvalue increases (Table Four) leads us to infer that changes in the composition and degree of the

13 Also according to the UNCTAD Statistics Database, the share in world manufacturing exports for Eastern,Southern, and Southeastern Asia excluding China was 16% in 1995, 20% in 2000 and 23% in 2005.


7/47


8/47

8

Table Five: Sector Unit Value Representation in Total Dollar Value per Exported Unit, by SITC Code

(in US Dollars)

Totaltrade

0Foodand liveanimals

2Crudematerials,inedible,exceptfuels

3

Mineralfuels,lubricantsandrelatedmaterials

4

Animalandvegetableoils, fatsandwaxes

5Chemicalsandrelatedproducts,n.e.s.

6Manufacturedgoods

7Machineryandtransportequipment

8 Miscellaneoumanufacturarticles

1994 0.51 0.04 0.02 0.02 0.00 0.03 0.10 0.09 0.2

1995 0.52 0.04 0.02 0.02 0.00 0.03 0.11 0.11 0.

1996 0.51 0.03 0.01 0.02 0.00 0.03 0.10 0.12 0.

1997 0.47 0.03 0.01 0.02 0.00 0.03 0.09 0.11 0.

1998 0.47 0.03 0.01 0.01 0.00 0.03 0.08 0.13 0.

1999 0.46 0.02 0.01 0.01 0.00 0.02 0.08 0.14 0.

2000 0.46 0.02 0.01 0.01 0.00 0.02 0.08 0.15 0.

2001 0.45 0.02 0.01 0.01 0.00 0.02 0.07 0.16 0.

2002 0.46 0.02 0.01 0.01 0.00 0.02 0.08 0.18 0.

2003 0.56 0.02 0.01 0.01 0.00 0.03 0.09 0.24 0.

2004 0.66 0.02 0.01 0.02 0.00 0.03 0.11 0.30 0.

Source: OECD Data and authors calculations

Further, the change in unit export values does not reflect or correspond to any large increase in

the wage bill. This can be seen in Table Six, which illustrates wages per unit output. We do not

see a marked decrease in this ratio, which would have indicated that a price increase is related to

a sudden increase in wage claims. Rather, wage costs appear to hold steady from 1997 onward.

Table Six: Wage to Output Ratio(in Percentages)

Year 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 200Wage

Bill 13% 12% 11% 11% 11% 11% 11% 11% 11% 11% 1

Source: China Data Center and authors calculations

The final issue is, to what degree could shifts in the composition of Chinese exports acrossnarrowly defined (three-digit) industrial categories within manufacturing account for theapparent rise in unit values? Table Twenty-Two in the appendix reports the results of adisaggregation exercise, aimed at isolating those categories with the largest increase in export


9/47

9

share due to apparent changes in unit value. We find that compositional shifts do occur, but theydo not appear to be very large in relation to the overall increase in reported unit value. And theirinterpretation is ambiguous. Interestingly, the largest changes are substantially concentrated incapital goods sectors such as machinery and equipment precisely those sectors where qualitychanges, quantity increases and disguised capital inflows would be hardest to disentangle. We

are left with no clear resolution on this topic, While it is possible that China suddenly andrapidly upgraded the quality of its machinery exports after 2002, it is also possible that thoseseeking to disguise capital inflow would tend to choose these same sectors as being the safestchannel for such activity. Detailed forensic work, case studies of technical change in the relevantsectors, and insider accounts would appear to be necessary to resolve the issue.

We infer that Chinas export figures overall, as well as in the important manufacturing sectors,are very probably overstated. By how much?

We now examine the extent to which funds may have entered China via this illicit route. A clueto the phenomenon at hand may possibly be found in the percentage change in gross capitalformation (Table Seven). This figure increases sharply in the post 2002 years. This is the result

of an enormous increase in the construction of fixed assets such as plant and equipment, officesand housing.

Table Seven: Gross Capital Formation(In Current US Dollars or Percentages where indicated)

GDP (billionsof current US$)

Gross capitalformation

(billions ofcurrent US$)

Percentagechange in

gross capitalformation

Percentagechange ingross capitalformationadjusted for

min hotmoney inflows

Percentagechange ingross capitalformationadjusted for

max hotmoney inflows

1996 856.1 346.2

1997 952.7 361.5 4% 4% 4%

1998 1,019.5 378.2 5% 5% 5%

1999 1,083.3 398.0 5% 5% 5%

2000 1,198.5 420.9 6% 6% 6%

2001 1,324.8 480.5 14% 14% 14%

2002 1,453.8 550.5 15% 15% 15%

2003 1,641.0 676.1 23% 18% 7%

2004 1,931.7 835.7 24% 19% 8%2005 2,243.9 971.0 16% 12% -2%

2006 2,668.1 1,085.8 12% 7% -11%


10/47

10

Source: WDI and authors calculations

The increase in gross capital formation reflects, in other words, the construction boom that is

everywhere visible in urban China. Gross capital formation increased by more than 60

percentage points from 2003 to 2006. What is more, to take a specific instance, the Beijing real

estate industry operating income and profit moves sharply from negative to positive numbers in

2003, a dramatic increase. This is a very significant change.

Table Eight: Beijing Real Estate Statistics(in Million Yuan)

Real Estate Industry

Operating Income


Total Profits

Investment in

Office Buildings

Commercial

Buildings Sold

2000 -1862 -1303 4521.9 424.84

2001 -1046 -215.3 7199.3 1245.8

2002 -1026 -587.1 9732.6 2595.3

2003 895.9 1743.3 14275 5177.9

2004 8661.1 10701 18789 5883.4

2005 6184.4 8131 19617 12085

2006 11053 14959 21674 16256

Source: CEIC

The question, then, is: how much of this increase might be accounted for by capital inflow?

We believe the answer could be: much of it. If we assume, conservatively, that 30% of theincrease in export unit value is due to disguised capital inflows, we find that China is running amuch smaller trade surplus. In this case, we estimate that the total disguised capital inflows intothe export account were USD $23 billion in 2003, $54 billion in 2004, $95 billion in 2005, and$157 billion in 2006. This accounts for much of the rise in fixed investments as a share of GDPthat had occurred up to that point. At the other extreme, based in part on unit price indicesreported by importers of Chinese manufactures, it would not be unreasonable to argue that therewas no increase in real unit export values after 2002. The disguised capital inflows would

amount to $87 billion in 2003, $199 billion in 2004, $347 billion in 2005 and $529 billion in2006. In that case, Chinas 2006 actual current account deficit would amount to $425 billion,and the cumulative deficit since 2003, disguised by capital inflow, would amount to $847 billion.


11/47

11

Table Nine: Capital inflows under varying assumptions of unit value increase

(in USD)

Percentage ofIncrease from2002 2003 2004 2005 2006

30% 26.0 59.6 104.0. 158.6

60% 52.0 119.2 208.0 317.2

90% 77.9 178.9 312.1 475.8

100% 86.6 198.7 346.7 528.6

Source: OECD, WDI and CEIC Data and authors calculations

Table Ten: Effect of estimated capital inflows disguised as current account on Balance of Trade(in Current billions of US Dollars)

Exports ofGoods and

Services

Exports ofGoods andServices

Minus HotMoney

Inflows

Difference(Max Hot

Money

Inflows)

Imports ofGoods and

Services

Balance of

Trade

Balance ofTrade

Adjustedfor

MinimumValue

(30%) ofHot Money

Inflows

Balance ofTrade

Adjustedfor

MaximumValue

(100%) ofHot Money

Inflows

1998 207 207 0 164 44 44 44

1999 221 221 0 190, 31 31 31

2000 280 280 0 251 29 29 29

2001 299 299 0 271 28 28 28

2002 365 365 0 328 37 37 37

2003 485 398 87 449 36 10 -51

2004 656 457 199 607 49 -10 -149

2005 837 490 347 712 125 21 -222

2006 981 452 529 878 103 -55 -425

Source: WDI and authors calculations


12/47

12

Caveats and Qualifications

The assumptions used in our calculations are subject to several qualifications and reservations.First, we assumed that growth in unit value data from the CEIC database is transferable to OECDdata. Since SITC category and overall quantity totals are not given in the CEIC database (only

subcategory values are available), the translation may not be entirely accurate for 2005-6. Whenwe asked the OECD statistics division why sub-category totals were smaller than SITC categoryand overall totals, they replied that this was due to confidential reasons on the part of theChinese government. We incorporated only values and quantities for CEIC categories that hadboth types of data available, and believe that this may underestimate the unit value increases.Therefore, on this account, we erred on the side of understating capital inflows.

Second, there are issues with Chinese reported statistics as noted in other literature. There areproblems with GDP, particularly with the overstatement of GDP for political purposes, as well asproblems with trade statistics, due to Hong Kong re-exports16. Although this is problematic indetermining exact numbers, the phenomenon of inflated export figures is more or less traceable

since the last major shift in statistics occurred in 1998, when the National Bureau of Statisticsbegan to use sample survey estimates of small scale industry, affecting the calculation of GDP(Naughton 2007, p. 141). A smaller shift in statistical classification has occurred in the past fewyears, when several export categories were broken into sub-categories, while some werediscontinued. However, this did not affect trade statistics within the larger SITC categories. .

Changes in the Financial Environment

Assuming that China has, in fact, experienced major capital inflow disguised as export earnings,why did it happen? In part, we believe, because changes in Chinas regulatory environmentmade it possible.

In 2003, there were several changes in Chinas financial sector which made the environmentmore favorable to capital inflows. The interest rate began to look more attractive vis--vis thedollar, while the NDF premium began to decrease, indicating expectations of yuan appreciationagainst the dollar. Tables Eleven and Twelve illustrate these trends.

16 Green writes tha t the US exag ge rate s value-ad de d in Hong Kong as around 25% of C hina s

goo ds va lue, wh ile Ch ina tends to und ersta te the se va lues. He be lieve s the US-China defic it

ma y be the a verage of the tw o reco rds. In any c ase, China s understa tem ent of Hong Kong re-

expo rts has not c hange d o ver time , so d oes not a ffec t the gene ral unit value trend .


13/47

13

Table Eleven: RMB Less Dollar Yields(In Basis Points)

1998 1999 2000 2001 2002 2003 2004 2005 2006

Avg 3-monthChineseRepo less USTreasury 1.96 -1.17 -3.4 -0.83 0.54 1.59 1.35 -1.44 -2.41

Avg 3-monthCHIBORless USDLIBOR 2.23 0.95 -2.46 0.03 1.6 1.66 1.71 -0.77 -2.57

Source: CEIC, US Treasury Statistics, British Bankers Association

Table Twelve: Non-Deliverable Forward Premium(Percentage of Spot)

Source: Ma and McCauley (2007, p. 16)


14/47

14

The rise in unit export values also occurred in conjunction with an important change in the rulesgoverning the holding of dollars inside China. In October 2002, the central government gavepermission for all companies to hold foreign exchange accounts. Controls over foreign exchangepurchases were relaxed for many businesses, including exporters, while the ability to openforeign exchange accounts was extended to firms outside bonded zones (Lehmanbrown.com,

2002). The goal of this measure was to liberalize the current account, facilitating trade andreducing the state presence in credit markets. Not surprisingly, we see, in Table Thirteen, thatforeign exchange transactions within China increased tremendously beginning in 2003.

Table Thirteen: Foreign Exchange Transactions within China

(in 100 Million Units)

OverallTurnover(in USD)

USDTradingVolume

HKDTradingVolume

JPYTradingVolume

EUROTradingVolume

2001 750.3 741.3 30.6 613.9 N/A

2002 971.9 951.1 108.8 730.8 1.1

2003 1511.3 1478.2 186.3 761.6 3.0

2004 2090.4 2044.1 244.9 1349.6 1.9

Source: Peoples Bank of China

Thus, the regulatory and investment environment was ripe for injecting capital inflows into

China. Using the trade account to bring in capital was relatively simple over this period.

Exporting companies simply had to overbill exports, and foreign exchange could be transferred

into the companies bank accounts.

The recent crackdown on short-term foreign exchange accounts, and the punishment of bothforeign and domestic banks for the violation of exchanging currency outside of controls, hasrevealed how loose controls over foreign exchange accounts had become. Further evidencecomes from the recent exposure and punishment of a large underground bank headquartered inShenzhen, which exchanged foreign currency and maintained foreign exchange accounts. All ofthese measures are attempts by the central government to curb hot money inflows and illegalforeign exchange transactions, in order to maintain better control over the current account

17.

Based on CEIC data available thus far this year, the unit value of exports has virtually ended itsupward movement, and perhaps the disguised inflow of capital has now come to an end.

Part of the increase, too, may stem from over-billing exports to receive additional Value-AddedTax 18(VAT) rebates after the January 2002 legislation loosened restrictions over VAT rebates.However, in our calculation, we do not see a large unit price increase for the year 2002, whichwould indicate that VAT abuses due to the legislation have not been very large.

17 In addition, the real appreciation of the renminbi in terms of the dollar in December 2006 signals a change in thedesirability of purchasing RMB with dollars.18 VAT rates range from 5-17%. The standard VAT rate is 17%.


15/47

15

What is more, accession to the World Trade Organization affected the trade climate after 2001,but the process of trade and capital control reform continued to be gradual. We believe, then,that much of the export unit price increase is due to overstated values that hide hot moneyinflows into Chinas real estate and other asset markets.

Implications of capital inflow via Chinas trade account

It appears that China ran a true current account surplus much smaller than reported, and mayhave actually run a trade deficit from 2003 to 2006. This conclusion depends in part on anassumption that Chinas import statistics are accurate. Though in the past the over-invoicing ofimports may have served to mask capital outflow from China, we have seen no evidence tosuggest similar distortions in Chinas import accounts at the present time. Indeed there would belittle reason for it: hot money flows where markets are hot, as they unquestionably have been inChina, and where a currency is widely expected to appreciate. This too was the Chinese case.

There are several implications of large net capital inflows disguised as exports. At a glance, trade

statistics, and any calculations which use export or net export values, will require correction

19

.This includes both GDP and the growth rate, figures envied by most other developed anddeveloping nations alike. Also suspect are the recent large profit increases reported by manyChinese firms, which could be an artifact of laundering exaggerated export earnings.

Given that we find a much smaller current account surplus, or even a deficit, repeated calls bythe United States for appreciation of the RMB based on evidence of an exploding trade surpluslose force. An appreciation of the RMB would increase the exchange value of the hot moneyinvested in China from 2003 to 2006, and the most interested parties may be speculators including some with political connections in the United States -- who have engaged in illicitinvestment in China through fictitious trade. This reduces much of the demand for RMBappreciation to an interest in validating a currency speculation.

The nature of Chinas external financial balance would also change. As an investment on behalfof foreign interests, capital inflow places foreign claims on domestic assets. Although Chinesedomestic savings and investment are high relative to other countries, domestic savings andinvestment will be seen as much lower than they have seemed, while foreign investment will beseen as much higher, once capital inflows through trade are correctly accounted for,

Conclusion

We believe that simple macroeconomic evidence points quite strongly to a significantoverstatement of Chinas exports, masking an equivalent capital inflow. This inflow ispotentially large enough to put Chinas actual current account into deficit, greatly weakening thecase for appreciation of the RMB. It also suggests that other aspects of the widely-held view ofrecent Chinese economic performance, including the profits boom in Chinese enterprises and thegrowth rate of the economy overall, should perhaps be re-examined for evidence of the role ofcapital inflow in distorting both the statistics and the underlying economy.

19 Some studies have shown that Chinas GDP statistics are overstated for reasons other than errors in the tradeaccount. The overstatement has taken place because some firm managers exaggerate output.


16/47

16

References

1. Anderlini, Jamil. China Hits Out Over Hot Money, Financial Times, June 27, 2007.2. Chinadaily.com. May 19, 2004, http://www.chinadaily.com.cn/english/doc/2004-

05/19/content_332004.htm

3.

Glick, Reuven and Hutchison, Michael. Capital Controls and Exchange Rate Instabilityin Developing Economies, Pacific Basin Working Paper Series, Center for Pacific BasinMonetary and Economic Studies, Economic Research Department, Federal Reserve Bankof San Francisco, 2000.

4. Green, Stephen. On the Ground in Asia, Standard Chartered Bank, Shanghai, April 13,2006.

5. Green, Stephen. On the Ground in Asia, Standard Chartered Bank, Shanghai, May 17,2007

6. Lehmanbrown.com, 2002. http://www.lehmanbrown.com/FAQ/FAQ-Forex/2.htm7. Ma, Guonan and Robert N. McCauley. Do Chinas Capital Controls Still Bind?

Implications for Monetary Autonomy and Capital Liberalization, BIS Working Paper

No. 233, www.bis.org.8. Naughton, Barry. An Economic Bubble? Chinese Policy Adapts to Rapidly ChangingConditions, China Leadership Monitor, No. 9, 2003.

9. Naughton, Barry. The Chinese Economy, MIT Press: Cambridge, MA 2007.10.Peoples Bank of China. Peoples Bank of China Adjusted Foreign Exchange

Administration Policy towards Individuals, January 23, 2007,http://www.pbc.gov.cn/english//detail.asp?col=6400&ID=791.

11.Prasad, Eswar and Wei, Shang-Jin. The Chinese Approach to Capital Inflows: Patternsand Possible Explanations, IMF Working Paper, 2005.

12.Setser, Brad and Rosenblatt, Casson. RGE China Reserve Watch, February 2006.13.Wei, Shang-Jin and Zhang, Zhiwei. Collateral Damage: Exchange Controls and

International Trade, NBER Working Paper No. 13020, April 2007.14.World Bank. World Development Indicators Online.15.Xin, Zhiming. Inflows of Hot Money to be Curbed, chinadaily, June 27, 2007,

http://www.chinadaily.com.cn/china/2007-06/27/content_903359.htm.16.Zhang, Ran. Hot Money Influx is Cooling Down, July 13, 2007,

http://www.chinadaily.com.cn/china/2007-07/13/content_5433984.htm.17.Zhong, Weike. Control of Exports, China Chemical Reporter, June 16, 2007,

http://goliath.ecnext.com/coms2/summary_0199-6672809_ITM.


17/47

17

Appendix

Table Fourteen: Change in Unit Price, Year on Year, by Sector

(in Percentages)

Totaltrade


2Crudematerials,inedible,exceptfuels

3Mineralfuels,lubricantsandrelatedmaterials

4Animalandvegetableoils, fatsandwaxes

5Chemicalsandrelatedproducts,n.e.s.

6Manufacturedgoods


8Miscellaneomanufacturearticles

1995 3% -17% -11% 10% -23% 22% 17% 20% -8

1996 -3% -2% -12% 6% -21% -7% -16% 7% -1

1997 -7% -17% -20% -10% 32% -12% -7% -5% -4

1998 1% -4% -16% -26% -52% 1% -5% 15% 0

1999 -2% -9% 2% -17% -61% -8% -6% 8% -5

2000 0% -8% -11% 32% -31% -9% 0% 10% -7

2001 -3% -5% -15% -2% -13% 1% -6% 5% -8

2002 3% -4% -12% -16% -26% -4% 1% 12% -3

2003 21% 8% 3% 18% 6% 15% 17% 33% 12

2004 18% -6% 2% 14% 12% 18% 28% 25% 9


Table Fifteen: Export Unit Values of ASEAN Countries20

(in US Dollars)

1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

Indonesia 109 104 81 65 100 90 96 103 120 81

Philippines 146 134 105 121 100 84 77 79 77 N/A

Singapore 120 112 97 96 100 93 91 90 93 96

Thailand 127 122 107 102 100 102 97 105 118 130

Source: UN Comtrade Yearbook 2005

20 We show all available values for ASEAN countries.


18/47

18

Table Sixteen: U.S Import Price Indices (from World21) by Commodity Category

Foodandliveanimals

Beveragesandtobacco

Crudematerials,inedible,exceptfuels

Mineralfuels,lubricantsandrelatedmaterials

Chemicalsandrelatedproducts,n.e.s.

Manu-facturedgoods

Machineryandtransportequipment

Misc.manu-facturedarticles

Com-moditiesandtran-sactions,n.e.s.

1991 96 86 82 63 96 91 105 98

1992 93 87 86 60 97 90 106 100 1221993 95 87 91 48 96 90 108 101 136

1994 111 88 102 56 104 98 110 102 137

1995 105 91 111 59 106 104 112 104 1381996 103 93 106 80 105 98 110 103 133

1997 108 97 103 61 101 99 105 103 110

1998 103 98 92 38 97 94 102 101 105

1999 103 100 101 83 98 97 100 101 104

2000 100 101 97 106 101 100 99 100 952001 95 103 90 61 97 92 98 99 99

2002 99 103 95 95 98 94 96 99 114

2003 101 104 108 108 101 98 95 100 1392004 112 107 126 141 110 110 95 101 157

2005 117 109 134 202 115 114 94 101 171

Source: BLS Data22

21 Statistics on US imports from China exist (they are collected by BLS) but begin only in 2004, which isinsufficient for our purposes, but even then they show that since 2004, the import price index from China to the UShas held steady or is slightly declining.22 Category 4 was not available, and Category 971 out of 9 was the only available category. Also, some monthswere missing, so we used data from month 12.


19/47

19

Table Seventeen: EU15 Unit Value Index (2000 = 100), Imports from China (excl HK)

Source: Eurostat Online and authors calculations23

23 Yearly data was calculated by averaging monthly data

Foodandlive

animalschieflyforfood

Beverages andtobacco

Crude

materials,inedibleexceptfuels

Mineralfuels,

lubricantsandrelatedmaterials

Animaland

vegetableoils, fatsandwaxes

Chemical

s andrelatedproducts,n.e.s.

Manu-facturedgoods

Machineryandtransportequipment

Misc.manu-facturedarticles

TotaAllprod

1995 88 67 81 87 82 99 87 92 77

1996 92 75 85 96 99 101 89 92 81

1997 97 94 93 103 119 103 95 95 891998 96 97 94 102 125 98 92 91 88

1999 96 96 90 92 105 94 89 90 882000 100 100 100 100 100 100 100 100 100

2001 100 100 103 127 98 99 101 100 99

2002 98 92 91 112 83 91 94 97 93

2003 87 83 84 127 78 82 84 88 822004 82 75 90 261 84 78 83 88 78

2005 85 75 97 217 94 82 85 88 79

2006 94 77 97 187 90 83 89 89 85


20/47

20

Table Eighteen: Japanese Imports from World, Unit Price Index on Yen Basis,(2000 average=100)

Allcom-modities

Foodstuffs&feedstuffs Textiles

Metals&relatedproducts

Wood,lumber&relatedproducts

Petroleum,coal &natural gas

Chemicals& relatedproducts

Machinery&equipment

Othergoods

1991 118 122 116 125 110 99 103 143 1161992 111 120 110 114 115 88 94 139 110

1993 100 109 98 93 141 73 90 125 100

1994 94 107 99 95 125 62 95 117 971995 94 106 98 100 117 63 104 110 100

1996 103 119 106 102 131 80 102 115 107

1997 111 123 113 112 133 92 111 119 113

1998 105 123 118 108 111 73 108 122 1171999 96 108 107 94 107 71 98 108 105

2000 100 100 100 100 100 100 100 100 100

2001 102 110 103 101 104 107 105 97 106

2002 101 113 103 100 107 105 105 93 1072003 100 116 100 102 104 112 110 85 103

2004 104 124 99 125 111 124 115 80 104

2005 118 127 100 153 113 172 124 78 109

2006 137 135 105 216 132 216 139 81 120

Source: Statistics Bureau, Ministry of Internal Affairs, Japan and authors calculations24

24 Yearly data was calculated by averaging monthly data


21/47

21

Table Nineteen: Change in Unit Quantity, Year on Year, by Sector

(in Percentages)

Totaltrade


2

Crudematerials,inedible,exceptfuels

3Mineral

fuels,lubricantsandrelatedmaterials

4Animal

andvegetableoils, fatsandwaxes

5

Chemicalsandrelatedproducts,n.e.s.

6Manufacturedgoods


8Miscellaneomanufacturearticles

1995 19% -56% 0% 22% -20% 26% 79% 25% 13

1996 5% 8% -7% -1% -7% 0% -15% 13% 53

1997 30% 65% 17% 18% 71% 19% 17% 26% 70

1998 0% 3% -9% -3% -58% -4% -22% 17% 2

1999 9% -2% -7% -1% -63% 16% -9% 29% 8

2000 27% 45% 30% 35% 9% 19% 32% 31% 11

2001 10% -16% 35% 35% 15% 16% -1% 2% 22002 19% 40% 46% -5% -19% 10% 14% 30% 17

2003 11% 29% -31% 10% -29% 32% 19% 18% 16

2004 14% -42% -1% -10% 15% 22% 37% 32% 9



22/47

22

Table Twenty: Change in Total Reported Value by Sector

(in Percentages)

1996 1997 1998 1999 2000 2001 2002 2003 2004

Total trade 2% 21% 1% 6% 28% 7% 22% 35% 35%0 Food andlive animals 3% 8% -4% -1% 17% 4% 14% 20% 8%

1 Beveragesand tobacco -2% -22% -7% -21% -3% 17% 13% 4% 19%

2 Crudematerials,inedible,except fuels -7% 4% -16% 11% 14% -6% 6% 14% 16%

3 Mineralfuels,lubricants andrelatedmaterials 11% 18% -26% -10% 69% 7% 0% 32% 30%

4 Animal andvegetable oils,fats and waxes -17% 72% -53% -57% -12% -4% -12% 18% 29%

5 Chemicalsand relatedproducts, n.e.s. -2% 15% 1% 1% 17% 10% 15% 28% 35%

6Manufacturedgoods -12% 21% -6% 2% 28% 3% 21% 30% 46%

7 Machineryand transportequipment 12% 24% 15% 17% 40% 15% 34% 48% 43%

8 Miscmanufacturedarticles 4% 25% 0% 3% 19% 1% 16% 25% 24%

9Commoditiesandtransactions,n.e.s. -46% 93% -98% 3022% 154% 18% 15% 49% -14%



23/47

23

Table Twenty-One: Share in Value of Total Trade, by Sector

(in Percentages)

1996 1997 1998 1999 2000 2001 2002 2003 2004

0 Food and live animals 7% 6% 6% 5% 5% 5% 4% 4% 3%

1 Beverages andtobacco 1% 1% 1% 0% 0% 0% 0% 0% 0%

2 Crude materials,inedible, except fuels 3% 2% 2% 2% 2% 2% 1% 1% 1%3 Mineral fuels,lubricants and relatedmaterials 4% 4% 3% 2% 3% 3% 3% 3% 2%

4 Animal and vegetableoils, fats and waxes 0% 0% 0% 0% 0% 0% 0% 0% 0%

5 Chemicals and relatedproducts, n.e.s. 6% 6% 6% 5% 5% 5% 5% 4% 4%

6 Manufactured goods 19% 19% 18% 17% 17% 16% 16% 16% 17%

7 Machinery andtransport equipment 23% 24% 27% 30% 33% 36% 39% 43% 45%

8 Miscellaneousmanufactured articles 37% 38% 38% 37% 35% 33% 31% 29% 26%

9 Commodities andtransactions, n.e.s. 0% 0% 0% 0% 0% 0% 0% 0% 0%



24/47

24

Table Twenty-Two: Top Twenty Subcategories (within SITC 6 through 8) for Change in Share due to

Value Change25

SITC 6 through 8, 3-digit level category

Change in Share due to Value Change,

2002-4752 Automatic data processing machines, n.e.s. 11%

764 Telecommunication equipment, n.e.s.; &parts, n.e.s. 8%

759 Parts, accessories for machines of groups751, 752 5%

845 Articles of apparel, of textile fabrics, n.e.s. 3%

894 Baby carriages, toys, games & sporting goods 3%

776 Cathode valves & tubes; diodes; integratedcircuits 3%

763 Sound recorders or reproducers; televisionrecord. 3%

851 Footwear 3%

778 Electrical machinery & apparatus, n.e.s. 3%

842 Women's clothing, of textile fabrics 2%

821 Furniture & parts; bedding & similar stuffedfurni. 2%

841 Men's clothing of textile fabrics, not knitted 2%

893 Articles, n.e.s., of plastics 2%

772 Apparatus for electrical circuits; board,panels 2%

658 Made-up articles, of textile materials, n.e.s. 1%

699 Manufactures of base metal, n.e.s. 1%

848 Articles of apparel, clothing access.,excluding textile 1%

653 Fabrics, woven, of man-made fabrics 1%

871 Optical instruments & apparatus, n.e.s. 1%

771 Electric power machinery, and parts thereof 1%


25 To c alc ulate this, we look a t the c hang e in the unit va lue, relative to the a verage, from 2002-4

relative to the o rigina l percentag e o f total value in 2002. This ga ve us a p ercentage that

presents the unit-va lue c hang e c om po nent o f the shift in share tow ard the sec tor.


25/47

1

The Beijing Bubble: Inequality, Trade and Capital Inflow into China

James K. GalbraithProfessor

Lyndon B. Johnson School of Public AffairsThe University of Texas at AustinAustin, Texas

Sara HsuAdjunct Professor of Economics

St. Edwards UniversityAustin, Texas

Wenjie ZhangLBJ School of Public Affaits

The University of Texas at AustinAustin, Texas

May 31, 2008

The University of Texas Inequality Project

UTIP Working Paper 50

Abstract

This paper explores the relationships between inequality, trade and capital flows into China sincethe early 1990s. We show that the rise in inequality in China since 2000 has more to do with thespeculative activities associated with Chinas building boom, notably in Beijing, than with themassive growth in manufacturing employment and in Chinese exports since China joined theWTO in 2001. The paper also reports further research on the likelihood of large speculativeinflows of capital into China via the current account. An earlier argument for this phenomenon

based on inspection of apparent export unit values by sector did not withstand scrutiny in moredetailed data sets. Rather, it is the flow of profits from the export boom that has, most likely, fedthe speculative fires in the capital and elsewhere.


26/47

2

This paper explores the relationships between inequality, trade and capital flows into China since

the early 1990s and particularly in the first years of the present decade. We show that the rise in

inequality has more to do with the speculative activities associated with Chinas building boom,

notably in Beijing, than with the massive growth in manufacturing employment and in Chinese

exports since China joined the WTO in 2001. Nevertheless, it is the flow of profits from the

export boom that has, most likely, fed the speculative fires in the capital and elsewhere.

By all accounts, inequality rose rapidly in China beginning in the early 1990s (Riskin et al.,

2001). Measurements by Galbraith, Krytynskaia and Wang (2004) showed that much of the rise

in that decade could be attributed to the relative gains of just one province and two municipalities:

Guangdong, Shanghai and Beijing, and to the relative earnings gains of just three sectors:

transportation, utilities and banking. Major regional losers in relative terms included the

Northeast (Manchuria) and the Southwest (Sichuan); across sectors the major losers included

manufacturing, farming and trade.i

Figure One presents a broad overview of the evolution of pay inequality in China, overall and by

region and sector, through 2005. The method consists of calculating the contribution of each

sector within each province to the between-groups component of a Theil T statistic for the whole

country, and then aggregating the components by sectors and by provinces to achieve measures

of inequality between and within provinces. The figure shows that while during the 1990s

inequality between provinces and inequality within provinces (that is, between sectors) both rose,

in the 2000s the behavior of these two dimensions of inequality has diverged. Inequality


27/47

3

between provinces peaked early in the decade, and has actually declined since 2001. In contrast,

inequality within provinces continued to rise.

Figure 1. Inequality between and within provinces in China, 1987-2006.

The Overall Theil Inequality Index for China from 1987 to 2006

0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0.09

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

Year

Between-GroupsComponentof

Theil'sTStatistics

Overall Inequality Between Provinces Within Provinces

Source: China Statistical Yearbook and authors calculations

Figure Two breaks out the changing inter-regional dimensions of Chinese inequality in a stacked

bar graph. Each bar represents a year and each segment represents the contribution of a

province to overall inequality in that year. Each segment reflects both the population weight of

the province (measured by observed employment) and the ratio between average provincial

income and national average income. Contributions greater than zero indicate provinces with

mean incomes above the national average. Contributions below zero indicate provinces with

incomes below average. Overall inter-provincial inequality is measured by the sum of all the

elements in a given year; however the statistic is so constructed that longer bars represent higher

inequality and vice versa.


28/47

4

Figure 2. Contribution of provinces to inter-provincial inequality in China, 1987-2006.

-0.15000

-0.10000

-0.05000

0.00000

0.05000

0.10000

0.15000

0.20000

1989 1988 1987 1991 1990 1992 1993 1994 1996 1995 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006

Year

TheilElementsbetweenProvinces

Beijing Shanghai Guangdong Zhejiang Jiangsu Tianjin Tibet Qinghai Ningxia

Hainan Chongqing Inner Mongolia Xinjiang Liaoning Yunnan Gansu Guizhou Fujian

Guangxi Anhui Shanxi Jilin Shaanxi Hunan Jiangx i Sichuan Shandong

Heilongjiang Hebei Hubei Henan

Source: China Statistical Yearbook and authors calculations

The figure shows that the enormous relative contribution of Guangdong province to overall

inequality in China actually peaked as far back as 1994, while that of Shanghai reached its zenith

around 2000 or 2001. Despite their respective positions as the seat of Chinese export trade and

the financial center, both were regressing moderately toward mean income by 2005 -- as incomes

elsewhere rose. Uniquely among the big three, the relative contribution of Beijing continued to

rise, reflecting in part, no doubt, the acceleration of a program of urban reconstruction and a

speculative building boom in advance of the 2008 Olympics. The recent rise of a fourth

contender Zhejiang province rounds out the contrasting picture of convergence and


29/47

5

divergence as the great Chinese coastal development boom matures. able One presents some

evidence on trends in manufacturing employment across China during the early years of the new

millennium. The table shows that in most Chinese provinces manufacturing employment

declined from 2002 through 2006. But there are great exceptions: Guangdong, Zhejiang, Fujian,

Jiangsu and Shandong, where manufacturing employment rose by a cumulative total of 4.9

million jobs during these four years. All are deeply involved in Chinaintegration into world

markets following accession to the WTO. Their expansion offset a net decline in manufacturing

employment of 1.5 million jobs spread across the rest of the country, giving China as a whole a

net gain in manufacturing employment exceeding ten percent in that period. Or, in four years

these five Chinese provinces added manufacturing jobs equal to thirty-six percent of the

remaining manufacturing employment in the United States as of April, 2008.

Table 1. The total number of manufacturing workers by provinces (10000 persons)

Source: State Statistical Yearbook

The total number of manufacturing workers in the major export provinces (10000 persons)

Year/Region Jiangsu Fujian Shandong Guangdong Zhejiang Rest of country

2002 216 134 272 255 97 1933

2003 217 152 270 282 109 1869

2004 223 181 280 315 152 1810

2005 245 198 334 357 201 1762

2006 281 215 342 387 240 1786

It is obvious that these gains in manufacturing employment are closely tied to exports. After

rising at just over 10 percent per year, on average, from 1999 through 2001 (two years of boom

and one of recession in the US), Chinas exports started to surge in 2002. They rose 21 percent

that year, and then 35 percent in each of the two following years, before settling back to a

reported rate of 28 percent in 2005 and 27 percent in 2006. Overall the reported increase in


30/47

6

exports in dollar terms from 2002 to 2006 amounts to a staggering 264 percent. There is no

question that a large part of this is real, in the sense that reported quantities surged, alongside

manufacturing employment in the key exporting provinces.

It is interesting that, apart from the rise of Zhejiang, the post-2001 export boom in China had

little effect on inequality as measured between provinces. The explanation is however not far to

seek. Though manufacturing in China is a low-wage sector, in high-wage provinces pay in

manufacturing can be close to, or even slightly above, national average pay rates. Thus an

increase in the manufacturing share of employment would not necessarily increase overall

inequality in China: the contribution of a sector whose average pay is close to the national

average to overall inequality is necessarily small. This is sufficient to explain why strong growth

in export-oriented manufacturing employment need not have had a dramatic impact one way or

the other -- on the inequalities of Chinese society. In contrast, the much higher incomes in

banking, utilities, government and real estate in Beijing have a powerful effect on inequality;

there is little else in the country quite like them.

Table Two presents the Chinese current account, as officially reported. It may be considered in

light of one of the most basic principles of international macroeconomics, that the growth of

imports depends on the domestic growth rate, while that of exports depends on growth in

external markets. Thus when a developing country experiences a prolonged period of high

internal growth, it is normal for a trade deficit to emerge. This is especially likely if the country

in question is an importer of food and fuel, and if commodity prices are rising. Innumerable

cases can be cited; exceptions, per contra, are rare, and in the modern record largely confined to


31/47


32/47

8

We have examined a number of explanations for this extraordinary turn of events. First, we

find no trace of any transformation in unit prices of Chinese exports. Information on the unit

prices of imports from China are maintained by European authorities, while the U.S. reports

price indices of imports in general. Nothing of consequence seems to have happened in either

dataset; indeed dollar prices of Chinese manufactures imported into Europe fell (not surprisingly,

given the rise of the euro against the dollar).

Although data we analyzed in a previous working paper (Galbraith, Hsu and Li, 2007) suggested

that there was a large (and suspicious) increase in reported unit values of Chinese exports after

2002, further research has deflated this conclusion. Our original hypothesis was that quantity

units reported by major product category in OECD summary data could be assumed to be

reasonably consistent over short time periods, permitting us to use aggregated, heterogeneous

quantities as a rough index of actual shipments. Inspection of the underlying data tables from

Comtrade reveals that large changes in reported units did occur (in some instances shifting from

actual units to thousands of the same units); thus in most (though not all) categories the

hypothesis of extraordinary changes in unit value cannot be sustained.

There have been modest shifts in the composition of exports toward higher-valued goods,

notably an increase of about three percentage points per year in the export share of the machinery

and transport equipment sector. But this increase had been going on for a long time, and the

gains after 2002 are not out of line with past experience. So while China is always in the process

of upgrading its manufactured exports, the major push behind the post 2001 boom has been


33/47

9

expansion in reported shipments (quantities) rather than in the value-added associated with

particular units (price or quality change).

A fourth possibility appears to have better traction: the decline of the dollar. China exports to

both Europe and the US, and while the relative stability of the dollar/RMB exchange rate assures

that the dollar value of Chinese exports to the US does not fluctuate with the dollar itself, this is

not true for Chinese exports elsewhere. Notably, if prices in the final markets do not change, then

the euros rise would automatically generate larger per-unit dollar earnings for China. But the

same effect would work on imports from outside the dollar zone, so it is difficult to see how this

artifact of the reference currency would strongly affect the rise in Chinas tradesurplus.v

A fifth possibility concerns the processing trade, a large share of Chinas manufactured exports.

China could be importing increasingly high-value goods (from, say, Japan) in order to finish

them and export them again. But if this were the case, then reported unit values of Chinese

imports in manufacturing would also be increasing, and so would the share of the processing

trade in total trade. Neither of these things appears to have occurred. Although there is a slight

progression in import prices in manufacturing from 2001 onward, no dramatic increase is

observed. Moreover, processing trade accounts for about 55 percent of Chinese exports, and that

figure remained stable after 2001.vi If the rise in total export values were due mainly to rising

unit value of processed goods, the share of the processing trade in total exports should have risen.

From the remarkable boom in manufacturing employment, coupled to the increase in unit sales,

it appears plain that after 2001 Chinas exporters took full advantage of their position as a WTO-


34/47

10

compliant country, and greatly multiplied their efforts and their results. This calls to mind a

comment in Galbraith (2006), following a discussion of the difficulties of making a profit on the

manufacture of wage goods for the home market, given the hyper-competitive climate for light

industry in China:

Is there any way for the Chinese manufacturing firm to turn a profit? Yes: the

obvious alternative to selling on the domestic market is to export. And export

prices, even those paid at wholesale must be many times those obtained at home.

It would not be surprising, therefore, if an export boom should lead to a profits boom, followed

by the speculative concentration of profit incomes in, for example, Beijing real estate. This

would appear to be the fundamental mechanism of rising inequality in China in the post-WTO

environment.

But if something can be done, it can also be overdone. Since China still maintains capital

controls, perhaps Chinese exporters have been over-reporting exports to the Chinese authorities,

for the purpose of bringing foreign capital into the country? Perhaps they have been over-

invoicing the exports they actually made? Or perhaps they have been, even more simply,

reporting exports to the authorities that were never made at all? The next section of this paper

considers this possibility, which has been discussed at least to some extent by Chinese officials.

There are straightforward reasons why it would be in the interest of Chinese firms to behave this

way, if they could get away with it. The incentive stems from Chinas property and stock market


35/47

11

booms, and from two regulatory facts: the continued enforcement of controls over capital inflows

per se in China, and the legalization, in late 2002, of unlimited foreign currency accounts held in

China by Chinese firms. The simple solution from the firms point of view in this situation

would be regulatory arbitrage: to launder the capital inflow through the current account.

At this point, we are unable to present estimates of the extent to which disguised capital inflow

may be occurring; as noted the preliminary estimates in earlier work already cited did not

withstand further scrutiny. Detailed forensic work, case studies of the relevant sectors, shipping

data, measures of unit imports from the advanced countries, and insider accounts would now

appear to be necessary to establish whether regulatory arbitrage is really a large issue. We

therefore restrict ourselves, for now, to examining the enabling conditions.

Changes i n the Financial Environment

In 2003, there were several changes in Chinas financial sector which made the environment

more favorable to capital inflows. The interest rate began to look more attractive vis--vis the

dollar, while the NDF premium began to decrease, indicating expectations of RMB appreciation

against the dollar (Ma and McCauley (2007, p. 16). Table Three illustrates the interest rate

trends.


36/47

12

Table 3: RMB Less Dollar Yields

(Percent)

Average 3-monthChinese Repo less US

Treasury Yield

Average 3-monthCHIBOR less USD

LIBOR

1998 1.96 2.23

1999 -1.17 0.95

2000 -3.4 -2.46

2001 -0.83 0.03

2002 0.54 1.6

2003 1.59 1.66

2004 1.35 1.712005 -1.44 -0.77

2006 -2.41 -2.57Source: CEIC, US Treasury Statistics, British Bankers Association

Further, in October 2002, the central government gave permission for all companies to hold

foreign exchange accounts. Controls over foreign exchange purchases were relaxed for many

businesses, including exporters, while the ability to open foreign exchange accounts was

extended to firms outside bonded zones (Lehmanbrown.com, 2002). The goal of this measure

was to liberalize the current account, facilitating trade and reducing the state presence in credit

markets. Not surprisingly, Table Four shows that foreign exchange transactions within China

increased tremendously beginning in 2003.
http://lehmanbrown.com/


37/47


38/47

14

However, in our calculation, we do not see a large unit price increase for the year 2002, which

would indicate that VAT abuses due to the legislation have not been very large.

Profit and Capital Inflows Into Speculative Sectors

We now examine the extent to which these funds both licit and otherwise -- may have

contributed to Chinas building boom and particularly to the Beijing Bubble. A clue to the

phenomenon at hand may possibly be found in the percentage change in gross capital formation.

This figure increases sharply in the post 2002 years, while the share of capital formation in GDP

rises by seven percentage points between 2001 and 2004. This is the result of an enormous

increase in the construction of fixed assets such as plant and equipment, offices and housing.

The increase in gross capital formation reflects the construction boom that is everywhere visible

in urban China. Table Five gives the basic information.


39/47

15

Table 5: Gross Capital Formation

GDP (billions of

current US$)

Gross capital

formation(billions ofcurrent US$)

Percentage

change ingrosscapital

formation

Share of

GrossCapital

Formation inGDP

(%) (%)1996 856.1 346.2 401997 952.7 361.5 4 381998 1,019.50 378.2 5 371999 1,083.30 398 5 372000 1,198.50 420.9 6 35

2001 1,324.80 480.5 14 362002 1,453.80 550.5 15 382003 1,641.00 676.1 23 412004 1,931.70 835.7 24 432005 2,243.90 971 16 432006 2,668.10 1,085.80 12 41

(Current US Dollars or Percent where indicated. Source: WDI and authors calculations)

An inflow of export profits, an increase in the profit share in total income, and any foreign

capital would need to show up as reported profits in Chinese industry not only directly but in

the sectors ultimately targeted by investment and speculation. This too we observe. To take a

specific instance, the Beijing real estate industry operating income and profit moves sharply from

negative to positive numbers in 2003, a dramatic increase. Table Six gives the data.


40/47

16

Table 6: Beijing Real Estate Statistics

(Million Yuan)


Operating Income


Total Profits

Investment in

Office Buildings

Commercial

Buildings Sold2000 -1862 -1303 4521.9 424.842001 -1046 -215.3 7199.3 1245.82002 -1026 -587.1 9732.6 2595.32003 895.9 1743.3 14275 5177.92004 8661.1 10701 18789 5883.42005 6184.4 8131 19617 120852006 11053 14959 21674 16256

Source: CEIC

Caveats and Qualifications

We note several further qualifications, arising from Chinese economic statistics as noted in other

literature. First of all, there are problems with Chinese GDP, particularly with the overstatement

of GDP for political purposes, and with the notorious stability of reported Chinese GDP growth

rates. There are well-known problems with the trade statistics, due to the treatment of re-exports

from Hong Kongx. There are also problems with achieving continuous measures of trade activity

over recent years, due to shifts in statistical classifications, e.g. as several export categories were

broken into sub-categories, while some were discontinued. However, this did not affect trade

statistics within the larger SITC categories.

Conclusion

While the rise in Chinese inequality seems to have slowed in the middle of the first decade of the

twenty-first century, a significant force for continued increases remained, associated with the


41/47

17

property boom and other speculative activities that concentrated on the national capital, Beijing,

during the period immediately before the 2008 Olympics. The most likely mechanism behind

the flow of funds into these sectors is a profits boom associated with the extraordinary increase

in Chinese exports that followed WTO accession in 2001. There is reason to suspect that some

additional speculative flows occurred by the device of laundering capital inflow through the

current account, but despite concentrated efforts we have no firm estimates to offer.


42/47

18

Appendix

Tables Seven, Eight and Nine present the sectors with the largest proportionate increases in

quantities and values exported. Interestingly, the largest quantity increases are in the metals

sectorsrolled steel, pipe, aluminum -- followed by certain electronics sectors. We have no way

of judging, a priori, whether an eleven-fold increase in flat-rolled steel exports in four years is

plausible or not; we only note that in this and other important sectors the numbers appear to be

remarkably high.

These sub-sectors show increases that are very large for such a short time frame, though they are not

beyond the bounds of possibility.


43/47

19

Table 7: Top 15 Categories (SITC 6 through 8) of Changes in Quantities Exported, 2002-6

SITC

Category

SITC Description Change in

Quantity, 2002-6

Dollar change in value

2002-6 (in millions)

673 Flat-rolled products of iron or non-alloy steel,

not clad, plated or coated

1126% 7220.2

7643 Transmission apparatus for radio-telephony,

radio-telegraphy, etc

653% 29707.5

679 Tubes, pipes and hollow profiles, and tube orpipe fittings of iron or steel

521% 6213.7

7722 Printed circuits 461% 5847

7764 Electronic integrated circuits and micro-

assemblies

414% 17148.6

7843 Other parts and accessor ies of the motor vehicles of 722 and 781-3

380% 7008.8

77812 Electric accumulators (storage batteries) 340% 3677.277121 Static converters (e.g., rectifiers) 272% -14721.9

6531 Fabrics, woven, of synthetic filament yarn

(not pile and chenille fabrics)

268% 7937.6

684 Aluminum 251% 5144.9

8943 Video games of a kind used with a televisionreceiver

242% 2795.9

747 Taps, cocks, valves, etc; pressure-reducing,

thermostatically control valves

229% 4186.5

7725 Electrical apparatus for switching, protecting

electrical circuits, for 1000 V

227% 3543.7

7649 Parts or accessories suitable for use solely orprincipally with apparatus of 76

217% 22511.2

763 Sound recorders or reproducers; television

image and sound recorders

215% 14825.7

Source: UN Comtrade


44/47

20

Table 8: Top 15 Categories (SITC 6 through 8) of Changes in Unit Values of Exported Goods,2002-6

SITC

Category

SITC Description Change in Unit

Value, 2002-6

Dollar change in value

2002-6 (in millions)

87193 Other opt ica l devices , appliances andinstruments

1201% 12379.2

793 Ships , boats ( inc luding hovercraft) and

floating structures

841% 6185.2

761 Television receivers 493% 10563.2

751 Office machines 270% 4637.1

752 Automat ic data process ing machines and

units thereof

220% 72885.1

7641 Electrical apparatus for line telephony or line

telegraphy

195% 6701.3

6531 Fabrics, woven, of synthetic filament yarn(not pile and chenille fabrics)

188% 7937.6

673 Flat-rolled products of iron or non-alloy steel,

not clad, plated or coated

186% 7220.2

671 Pig-iron, spiegeIeisen, sponge iron, iron orsteel granules and powders

179% 2026.1

684 Aluminium 171% 5144.9

773 Equipment for distributing electricity, nes 168% 5378.3

747 Taps, cocks, valves, etc; pressure-reducing,

thermostatically control valves

163% 4186.5

7649 Parts or accessories suitable for use solely or

principally with apparatus of 76

162% 22511.2

844 Women's or gir ls' outerwear, of text ilefabrics, knitted or crocheted

161% 7399

759 Parts and accessories (not covers, carrying

cases, etc) for machines of 751-52

157\% 20756.9

Source: UN Comtrade


45/47

21

Table 9: Top 15 Categories (SITC 6 through 8) of Changes in Volume-Weighted Unit Values ofExported Goods, 2002-6

SITC

Category

SITC Description Change

in

Volume-

weighted

Unit

Value,

2002-6

Dollar

change in

value

2002-6

(in

millions)

87193 Other optical devices, appliances and instruments 6265% 12379.2

673 Flat-rolled products of iron or non-alloy steel, not clad, plated

or coated

1304% 7220.2

793 Ships, boats (including hovercraft) and floating structures 1190% 6185.2

761 Television receivers 897% 10563.2679 Tubes, pipes and hollow profiles, and tube or pipe fittings of

iron or steel351% 6213.7

752 Automatic data processing machines and units thereof 341% 72885.1

6531 Fabrics, woven, of synthetic filament yarn (not pile and

chenille fabrics)

319% 7937.6

751 Office machines 285% 4637.1

684 Aluminium 248% 5144.9

7764 Electronic integrated circuits and micro-assemblies 213% 17148.6

7641 Electrical apparatus for line telephony or line telegraphy 208% 6701.3

671Pig-iron, spiegeIeisen, sponge iron, iron or steel granules and

powders 206% 2026.1

747Taps, cocks, valves, etc; pressure-reducing, thermostatically

control valves 205% 4186.5

7843Other parts and accessories of the motor vehicles of 722 and

781-3 204% 7008.8

773Equipment for distributing electricity, nes

191% 5378.3

Source: UN Comtrade


46/47

22

References:

James K. Galbraith, The Firm, Exports, the Banks and the Real Wage in China, A Note,Prepared for the meeting on Institutional Reform in China, Manchester, UK, August 8-9, 2006.

James K. Galbraith, Sara Hsu and Jianjun Li, Is China Really Running a Trade Surplus?, UTIPWorking Paper No. 45, December 30, 2007.

James K. Galbraith, Ludmila Krytynskaia and Qifei Wang, "The Experience of Rising Inequality inRussia and China during the Transition." European Journal of Comparative Economics, Vol 1, No. 1,

2004.

Carl Riskin, Zhao Renwei, Li Shi, editors. China's retreat from equality : income distributionand economic transition, Armonk, N.Y. : M.E. Sharpe, c2001.

Lehmanbrown.com, http://www.lehmanbrown.com/FAQ/FAQ-Forex/2.htm (2002)

Ma, Guonan and Robert N. McCauley. Do Chinas Capital Controls Still Bind? Implicationsfor Monetary Autonomy and Capital Liberalization, BIS Working Paper No. 233, www.bis.org.

Naughton, Barry. The Chinese Economy, MIT Press: Cambridge, MA 2007.

World Bank. World Development Indicators (WDI) Online.

About the Authors:

James K. Galbraith, LBJ School of Public Affairs, The University of Texas at Austin. Sara Hsu,

Trinity University, San Antonio, Texas. Wenjie Zhang, LBJ School of Public Affairs, TheUniversity of Texas at Austin.

i These results are drawn from data on pay and employment in the State Statistical Yearbook. They are consistentwith, but considerably more revealing than, surveys which have tended to characterize the growing gap in Chineseincomes as urban/rural or coast/interior.

iiThe IMFs world economic outlook pegs nominal GDP for China in 2006 at $2 trillion, in comparison to which

the official trade statistics look even larger. http://www.econstats.com/weo/index_glweo.htm. We have not tried to

unravel the discrepancy, except to say that all such comparisons are clearly open to skeptical appraisal.

vWere the reference currency switched to the euro, the rise in Chinas export earnings would appear lower, since

the countrys exports to the U.S., measured in euro, would have been sharply cut by the dollar devaluation. Similar

effects would apply on the import side: Chinas eurozone imports would not have risen so much, while its dollar-

zone imports would have risen considerably more.

vi http://www.mofcom.gov.cn/tongjiziliao/tongjiziliao.html
http://www.econstats.com/weo/index_glweo.htmhttp://www.econstats.com/weo/index_glweo.htmhttp://www.mofcom.gov.cn/tongjiziliao/tongjiziliao.htmlhttp://www.econstats.com/weo/index_glweo.htmhttp://www.bis.org/http://www.lehmanbrown.com/faq/faq-forex/2.htmhttp://lehmanbrown.com/


47/47

vi ii In addition, the real appreciation of the RMB in terms of the dollar in December 2006 signals a change in the

desirability of purchasing RMB with dollars.

ix VAT rates range from 5-17%. The standard VAT rate is 17%.

x

Green writes that the US exaggerates value-added in Hong Kong as around 25% of Chinas goods value, whileChina tends to understate these values. He believes the US-China deficit may be the average of the two records. In

any case, Chinas understatement of Hong Kong re-exports has not changed over time, so does not affect the general

unit value trend.
http://www.mofcom.gov.cn/tongjiziliao/tongjiziliao.htmlhttp://www.mofcom.gov.cn/tongjiziliao/tongjiziliao.html

galbraith china hot money

Documents