1 Antidumping and US-China Bilateral Intra-Industry Trade: An Empirical Analysis SHEN Guobing Associate Professor of World Economy & International Finance.

1

Antidumping and US-China Bilateral Intra-Industry Trade: An Empirical Analysis

SHEN GuobingAssociate Professor of World Economy & International Finance（ Institute of World Economy, Fudan University, 200433）

Homepage ： http://web.cenet.org.cn/web/economysoulE-mail : [email protected]

SCCS Economic Seminar, University of Nottingham, May 14, 2009

2

Antidumping and US-China Bilateral Intra-Industry Trade: An Empirical Analysis

Abstract: Focusing on the relation between antidumping (AD) and U.S.-China bilateral intra-industry trade (IIT), this paper uses the GL index and the binary logit model, and gets some valuable findings. In detail, the twenty goods groups, involved into the large friction tension of U.S. AD against China, have rather small indices of intra-industry trade during 1995~2006, from which the U.S. suffer quite large imbalance of trade deficit. These are the true reasons that the U.S. takes frequently AD investigations and definitive measures against China. In addition, this paper proves empirically that there is no positive relation between AD density and U.S.-China bilateral IIT. The conclusion is that the goods groups with rather small indices of U.S.-China bilateral IIT and quite large imbalance of trade deficit are easiest to cause the U.S. AD actions against China.

Keywords: Antidumping, Intra-industry Trade, GL index

3

1. Introduction Intra-industry trade refers to the international or inter-regional

trade of products and semi-finished products belonging to the same industry. In reality, the Sino-US product trade has formed the international differentiation and refinement with the increase of China’s export to US [1]. However, because of the differences of production scales, products and consumer preferences between China and US, intra-industry trade has developed continuously in some industries, such as the chemicals industry.

[1]According to the data reported by China in the UN trade statistic database, in 1995, the amount of China’s import from US was 16.12 billion USD, and the amount of China’s export to US was 24.73 billion USD; but in 2001, they had increased to 26.22 billion and 54.36 billion, respectively. After joining WTO, the amount of China’s import and export have increased quickly. In 2006, the amount of China’s import from US was 59.31 billion USD, and the amount of China’s export to US was 203.80 billion USD. According to the data reported by US, the amount of China’s import and export had increased much more. In 2006, the amount of US’s import from China was 305.78 billion USD, and the amount of US’s export to China was 55.22 billion USD. Cite in July, 2007.

4

According to the data reported by China in the UN trade statistic database, it can be calculated that the GL index [1] of Sino-US chemicals intra-industry trade had risen from 0.47 in 1995 to 0.81 in 2006. According to the data reported by US, the GL index had risen from 0.64 in 1995 to 0.94 in 2006. That is to say, Sino-US chemicals intra-industry trade is larger form the perspective of US, which means the competition between China and US in chemicals industry is becoming fiercer.

During 1995 to 2006, say, after the establishment of WTO, US or China AD investigations and AD definitive measures to the counterpart mainly concentrated on chemicals products.

[1]the GL index of intra-industry trade was introduced by Grubel and Lloyd (1975).

5

The questions are: is the density of antidumping positively related to Sino-US bilateral intra-industry trade? If not, what is the relation between antidumping and Sino-US bilateral intra-industry trade?

(1) Research on the intra-industry trade measure. Grubel and Lloyd (1975) originally introduced the intra-industry trade GL index:

Where Xj and Mj denote the export and import of the jth industry, GLj denotes the intra-industry trade of the jth industry to measure the balance or overlap of intra-industry trade.

( ) | |

( )j j j j

jj j

X M X MGL

X M

6

(2) Research on the industry trade and antidumping. Blonigen & Bown (2003)think that an industry is more likely

to file an AD petition the greater the import penetration and the lower its ‘exposure’ to retaliation. The industry is more ‘exposed’ to retaliation when the industry has significant exports to the same country it is petitioning against and when that country has AD policies in place.

Stevenson (2007): As Chinese imports of these sensitive products increase, it is possible that domestic industries will turn to anti-dumping where prices are low and there have been significant increases in imports.

7

2. The distribution of Chinese industrial products anti-dumped by US

In the early years of reform and opening-up (1978-1981), there was little antidumping lawsuit filed by US against China until 1980, which saw the first AD investigation against China, the menthol incident. But the AD investigations and AD definitive measures increased later on, while the distribution of involved industrial products expanded.

During 1978 to 1990, 7 industrial products were involved in US AD investigation against China: chemicals products, textile fabrics, food, base metals products, mineral manufactures, miscellaneous manufactures, and industrial/ electrical equipment. In 1991, transport equipment was involved. In 2000, mineral fuels were involved. In 2003, household appliance was involved. In 2004, paper articles were involved. As a result, 11 industrial products were involved in US AD investigation against China during 1978 to 2005.

8

In 1983, US gave the first verdict of AD definitive measure against China, the greige polyester cotton cloth incident and the cotton shop towels incident. During 1984 to 1996 US verdicts of AD definitive measure against China involved 7 industrial products. In 1997, transport equipment was involved. In 2001, mineral fuels were involved. In 2004, household electric appliance was involved. In 2005, paper was involved. As a result, 11 industrial products were involved in US AD definitive measure against China.

Specifically, chemicals products and base metals products were the most important industrial products of US AD investigations and AD definitive measures against China.

9

During 1978 to 2005, there are 38 AD investigation and 26 AD definitive measures to chemicals products, which occupied 32% of US AD investigation against China and 34% of US AD definitive measures against China, respectively. During the same period, there are 32 AD investigation and 22 AD definitive measures to base metals products , which occupied 27% of US AD investigation against China and 29% of US AD definitive measures against China, respectively. Furthermore, these numbers are increasing quickly.

In general, we can sort the industrial products which US anti-dumped against China during 1978 to 2005 as follows: chemicals products, base metals products, miscellaneous manufactures, food, mineral manufactures, industrial/ electrical equipment, transport equipment, textile fabrics, paper articles, mineral fuels and household appliance.

10

Table 1 The distribution of industrial products involved in US AD investigation and AD definitive measures against China

ItemNumber of US AD investigation against China Number of US AD definitive measures against China

1978~1991 1992~2005 1978~2005 1978~1991 1992~2005 1978~2005

chemicals products 12 26 38 8 18 26

textile fabrics 3 1 4 2 0 2

food 1 7 8 0 6 6

base metals products

7 25 32 4 18 22

mineral manufactures

3 2 5 3 2 5

miscellaneous manufactures

1 9 10 1 5 6

industrial/ electrical equipment

4 5 9 3 0 3

transport equipment

1 4 5 0 3 3

mineral fuels 0 2 2 0 1 1

household appliance

0 1 1 0 1 1

paper articles 0 3 3 0 2 2

Total 32 85 117 21 56 77

Source: http://people.brandeis.edu/~cbown/global_ad/;http://www.wto.org/english/tratop_e/adp_e/adp_e.htm#statistics

11

3. The main products and amount involved in US antidumping cases against China

We inspect the main products involved in US antidumping cases against China after the establishment of WTO in detail based on the amount involved in the cases. USITC (2006) statistics shows the main products involved in US antidumping cases against China whose amount is more than 100 million USD during the fiscal year 1995-2006. See Table 2 for reference.

Wooden bedroom furniture: involved amount 957.95 million USD, ruled 198.08% antidumping tax;

Carbon and certain alloy steel wire rod: involved amount 340.88 million USD, but preliminary ruling was rejected;

Certain frozen and canned warmwater shrimp and prawns: involved amount 295.30 million USD, ruled 112.81% antidumping tax;

Color television receivers: involved amount 271.11 million USD, ruled 78.45% antidumping tax;

Bicycles: involved amount 221.30 million USD, but ruled as no damage; Grid paper: involved amount 131.84 million USD, ruled 78.39%

antidumping tax; Ball bearings: involved amount 127.96 million USD, but ruled as no

damage; Polyethylene retail carrier bags: involved amount 125.72 million USD,

ruled 77.57% antidumping tax.

12

Table 2 Main products and amount involved in US antidumping cases against China (million USD)

Item filing date

Final ruling data

Ruling result

Involved amount

Item filing dateFinal ruling data

Ruling result

Involved amount

wooden bedroom furniture

2003.1 2004.1 tax 957.95non-frozen apple juice concentrate

1999.6 2000.6 tax 33.39

carbon and certain alloy steel wire rod

2005.1 2006.1initial ruling denied

340.88certain tissue paper products

2004.2 2005.3 tax 30.10

certain frozen and canned warm water shrimp and prawns

2004.1 2005.1 tax 295.30 honey 1994.1 1995.8agreement suspended

30.05

color television receivers

2003.5 2004.6 tax 271.11structural steel beams

2001.6 2002.6 no damage 27.07

bicycles 1995.4 1996.6 no damage 221.30 pure magnesium 2004.3 2005.4 tax 24.02

grid paper 2005.1 2006.9 tax 131.84 honey 2000.1 2001.11 tax 24.01

ball bearing 2002.2 2003.4 no damage 127.96citric acid and sodium citrate

1999.12 2000.2 no damage 22.54

polyethylene retail carrier bags

2003.6 2004.8 tax 125.72malleable iron pipe fittings

2002.11 2003.12 tax 20.40

circular-welded non-alloy steel pipe

2001.6 2002.7 no damage 68.18 steel wire rope 2000.3 2001.4 no damage 19.87

13

Table 2 Main products and amount involved in US antidumping cases against China (million USD)

Item filing date

Final ruling data

Ruling result

Involved amount

Itemfiling date

Final ruling data

Ruling result

Involved amount

canned mushrooms

1998.1 1999.2 tax 67.57

Refined brown aluminum oxide

2002.11 2003.11 tax 19.64

carbon steel standard pipe

1996.1 1997.1 tax 62.27synthetic indigo

1999.7 2000.6 tax 18.39

oil country tubular goods

2002.4 2002.5no damage

55.97diamond sawblades and parts thereof

2005.6 2006.7 no damage 18.24

automotive replacement glass windshields

2001.3 2002.4 tax 52.41cold-rolled carbon steel sheet

2001.1 2002.11 no damage 17.72

certain brake drums and rotors

1996.3 1997.4partially tax

51.47 paintbrushes 1999.8 1999.9 no damage 17.12

cold-rolled carbon steel products

1999.6 2000.7no damage

50.72 foundry coke 2000.9 2001.9 tax 12.94

crawfish tail meat 1996.9 1997.9 tax 35.70ironing tables and certain parts thereof

2003.7 2004.8 tax 12.44

pure magnesium 2000.1 2001.1 tax 35.46polyvinyl alcohol

2002.9 2003.1 tax 10.23

Resource: http://people.brandeis.edu/~cbown/global_ad/;http://www.usitc.gov/trade_remedy/Report-10-06-PUB.pdf.

14

In order to reveal the possible relation between antidumping and Sino-US bilateral intra-industry trade, we omit the products which were ruled as no damage or suspended the agreements to obtain the main products which US ruled antidumping tax on China:

Wooden bedroom furniture, certain frozen and canned warm water shrimp and prawns, color television receivers, grid paper, polyethylene retail carrier bags, canned mushrooms, carbon steel standard pipe, automotive replacement glass windshields, certain brake drums and rotors, crawfish tail meat, pure magnesium, non-frozen apple juice concentrate, certain tissue paper products, honey, malleable iron pipe fittings, refined brown aluminum oxide, synthetic indigo, foundry coke, ironing tables and certain parts thereof, polyvinyl alcohol, etc.

Then we can calculate the Sino-US bilateral intra-industry trade of the main products involved in antidumping cases to reveal the possible relation between antidumping and Sino-US bilateral intra-industry trade.

15

4. The bilateral intra-industry trade of the main products involved in US antidumping cases against China

Grubel and Lloyd (1975) originally introduced the intra-industry trade GL index:

Where Xj and Mj denote the export and import of the jth industry, GLj denotes the intra-industry trade of the jth industry to measure the balance or overlap of intra-industry trade. GL index is between 0 to 1. GL=0 means no intra-industry trade and totally unbalanced trade; GL=1 means totally intra-industry trade and bilateral balanced trade.

Hence, GL index is positively related with the balance of intra-industry trade and negatively related with the friction of bilateral trade.

( ) | |

( )j j j j

jj j

X M X MGL

X M

16

Now we calculate the intra-industry trade GL index of the products involved in US antidumping cases against China to consider which product has larger antidumping friction. See Table 3 for reference.

Table 3 The intra-industry GL indices of main products involved in US antidumping cases against China (see the next page)

17

Product/industry1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006

wooden bedroom furniture 0.007 0.043 0.007 0.007 0.001 0.001 0.002 0.001 0.002 0.002 0.001 0.001

certain frozen and canned warm water shrimp and prawns

0.16 0.18 0.10 0.17 0.26 0.06 0.01 0.03 0.04 0.01 0.25 0.19

color television receivers 0.04 0.11 0.30 0.27 0.07 0.09 0.08 0.03 0.02 0.02 0.02 0.02

grid paper 0.00 0.01 0.01 0.01 0.01 0.00 0.01 0.01 0.00 0.00 0.01 0.01

polyethylene retail carrier bags

0.02 0.01 0.01 0.01 0.02 0.02 0.01 0.02 0.02 0.04 0.03 0.02

canned mushrooms 0.00 0.00 0.01 0.01 0.08 0.00 0.00 0.01 0.01 0.00 0.00 0.00

carbon steel standard pipe 0.69 0.07 0.04 0.11 0.06 0.03 0.13 0.18 0.02 0.27 0.32 0.06

automotive replacement glass windshields

0.67 0.25 0.27 0.11 0.05 0.03 0.02 0.03 0.02 0.01 0.03 0.02

certain brake drums and rotors

0.03 0.02 0.02 0.02 0.05 0.12 0.03 0.06 0.05 0.03 0.06 0.10

crawfish tail meat / 0.08 0.00 0.00 0.00 0.00 0.29 0.00 0.23 0.16 0.32 0.35

non-frozen apple juice concentrate

0.02 0.05 0.00 0.01 0.02 0.02 0.06 0.00 0.00 0.00 0.00 0.00

certain tissue paper products

0.64 0.92 0.79 0.91 0.83 0.68 0.88 0.52 0.44 0.17 0.29 0.47

pure magnesium 0.20 0.00 0.00 0.03 0.01 0.01 0.00 0.02 0.00 0.03 0.06 0.24

honey 0.00 0.00 0.00 0.05 0.03 0.01 0.00 0.10 0.15 0.04 0.03 0.02

malleable iron pipe fittings 0.53 0.12 0.18 0.34 0.22 0.20 0.11 0.15 0.12 0.17 0.16 0.16

Refined brown aluminum oxide

0.10 0.08 0.75 0.35 0.43 0.17 0.40 0.14 0.83 0.46 0.25 0.26

synthetic indigo 0.08 0.06 0.03 0.08 0.15 0.17 0.17 0.16 0.20 0.21 0.22 0.19

foundry coke 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.01 0.02 0.04

ironing tables and certain parts thereof

/ 0.26 0.27 0.11 0.21 0.79 0.23 0.06 0.47 0.27 0.43 0.47

polyvinyl alcohol 0.15 0.27 0.13 0.10 0.15 0.13 0.19 0.95 0.83 0.93 0.90 0.99

Resource: UN trade statistic database http://comtrade.un.org/db, calculate by the data reported by US. Published in August, 2007.

18

From the table above we can see that the 20 products involved in US antidumping cases against China during 1995 to 2006 with large antidumping friction have small intra-industry trade GL index and make large trade deficit of US. That is the reason why there are continuous trade dispute launched by US against China.

It denies the positive relation between the density of antidumping and Sino-US bilateral intra-industry trade, but the conclusion should be testified econometrically.

19

5. The relation between antidumping and Sino-US bilateral intra-industry trade: empirical analysis Theoretically, there are two kinds of antidumping from the

perspective of Sino-US bilateral intra-industry trade: (1) When bilateral intra-industry trade is large, if US launches

antidumping, China will retaliate by antidumping, so the dominant strategy is both side do not launch antidumping.

(2) When bilateral intra-industry trade is small, the dominant side usually tends to launch antidumping, so US usually launches antidumping against China using its dominant strategy formed by trade deficit.

According to the result of game theory, antidumping usually occurs when bilateral intra-industry trade is small and one side has dominant strategy formed by trade deficit.

20

However, the intuitive description is limited to representation, so we choose the period 1995 to 2006 as a sample to inspect the internal relation between intra-industry trade and antidumping cases.

(1) Variables chosen and data description A. The products involved in antidumping cases We define a binary variable ADj to equal 1 if US

launched antidumping of the jth product against China and 0 if otherwise. See Table 4 to get variable ADj of all involved products.

21

B. The intra-industry trade of the involved products Because the GL index of a particular intra-industry trade

product has no distort of numerical offset induced by totaling, we can measure the intra-industry trade of the involved products by applying GL index directly.

Theoretically, antidumping usually occurs when bilateral intra-industry trade is small. GLjt denotes the jth product which US has intra-industry trade with China in year t.

C. The trade balance of involved products The intra-industry trade can only reflect the trade balance

between US and China, but cannot reflect whether US has antidumping dominant strategy, so we need to consider the trade deficit between US and China. Theoretically, the products with small intra-industry trade and US trade deficit tend to induce antidumping.

22

(2) Modeling and empirical results analysis Because the independent variable has only two choices: 0 when no

antidumping occurs, 1 when antidumping occurs, we use a binary choice model, and choose the main products involved in US antidumping cases against China to testify the proposition. Specifically, the binary choice model can write as:

Where ADjt* equals 0 or 1, GLjt denotes the Sino-US intra-industry

trade of the jth product in year t, NXjt denotes the US trade deficit of the jth product in year t, β0, β1, β2 are coefficients, μjt is error term.

According to the binary choice model, we choose 1996-2006 as a sample period and choose whether US anti-dumped the jth product against China as an independent variable. The model and the independent variables are chosen as above. When estimating the impact of Sino-US intra-industry trade of the involved products and trade deficit to induce antidumping, we choose between data of this period and data of lagged q period (q=1, 2, …) according to the significance of variables as well as the principle of minimized AIC, SC and maximized log-likelihood. The best test result of the binary choice model is as follows:

*0 1 2jt jt jt jtAD GL NX

23

Table 5 Regression result of the binary Logit model of main products involved in US antidumping cases against China

Dependent variable ： Whether US anti-dumped the jth product against China

Independent Variable

CoefficientStandard deviation

Z-statistics P-value

GL (-2) -1.026865 * 0.549317 -1.869348 0.0616

NX -0.113874 ** 0.050858 -2.239068 0.0252

Mean dependent var

0.521008 S.D. dependent var 0.500611

S.E. of regression 0.490900 Akaike info criterion (AIC) 1.352307

Sum squared resid 56.87201 Schwarz criterion (SC) 1.381485

Log likelihood -158.9245 Hannan-Quinn criter. 1.364066

Avg. log likelihood -0.667750

Obs with Dep=0 114 Total obs 238

Obs with Dep=1 124

Note: ** significant at the 5% significance level, * significant at the 10% significance level.Data resource: see “variables chosen and data description” for reference.

24

According to Table 5, we use Logit model defaulted by Eviews. Because the coefficient of the constant term is not significant, we omit it and obtain a binary Logit model without the constant term.

We also consider the possible lagged impact of Sino-US intra-industry trade of the involved products and trade deficit in the sample period to induce antidumping. Compare the significance of variables, minimized AIC, SC and maximized log-likelihood. Use the intra-industry trade GL index of involved products of 2-period lag. Use the trade deficit of this period.

At the significance levels of 5% and 10%, the two independent variables have statistically significant effects on the dependent variable.

25

At the 5% significance level, assuming other conditions remain unchanged, the probability that US launch antidumping to the jth product against China is negatively related with the US trade deficit with China. So it is testified that the products with trade deficit tend to induce the antidumping of US.

At the 10% significance level, assuming other conditions remain unchanged, the probability that US launch antidumping to the jth product against China is negatively related with the US intra-industry trade with China. However, the effect lags 2 periods.

So we deny the positive relation between the density of antidumping and Sino-US bilateral intra-industry trade empirically; furthermore, we testify that the small Sino-US bilateral intra-industry trade induce the frequent antidumping by US against China in the sample period.

26

6. Main conclusions and suggestions We obtain the following main conclusions from the empirical

analysis of antidumping and Sino-US bilateral intra-industry trade: We can sort the industrial products which US anti-dumped against

China as follows: chemicals, (cheap) manufactures of metals, commodities and transactions not classified elsewhere in SITC, food, mineral manufactures, machinery and electrical equipment, transport equipment, textile, paper, mineral fuels and household electric appliance.

In the fiscal year 1995-2006 the main products involved in US antidumping cases against China whose amount is more than 100 million USD were:wooden bedroom furniture, carbon and certain alloy steel wire rod, certain frozen and canned warmwater shrimp and prawns, color television receivers, bicycles, grid paper, ball bearings, polyethylene retail carrier bags. Other 26 products had involved amount of more than 10 million USD and less than 100 million USD. But a part of the products were ruled as no damage or suspended the agreements, such as , carbon and certain alloy steel wire rod, bicycles and ball bearings.

27

The 20 products involved in US antidumping cases against China during 1995 to 2006 with large antidumping friction have small intra-industry trade GL index and make large trade deficit of US. That is the reason why there are continuous trade dispute launched by US against China.

The regression of the binaryl Logit model shows that at the 5% significance levels, the probability that US launch antidumping to the jth product against China is negatively related with the US trade deficit with China; at the 10% significance levels, the probability that US launch antidumping to the jth product against China is negatively related with the US intra-industry trade with China. So we deny the positive relation between the density of antidumping and Sino-US bilateral intra-industry trade empirically.

Conclusion: the products with small bilateral intra-industry trade and US trade deficit tend to induce US antidumping against China.

28

Suggestions:

To increase the quality and value-added of the products exported to US by technology transfer and industry upgrading. Increase import and export of the same products at the same time to increase the Sino-US bilateral intra-industry trade so as to reduce the risk of antidumping partially.

To improve Sino-US merchandise trade structure by taxing on the products produced by resource-depleting or environment-polluting enterprises. Reduce the export of the products to US to reduce the huge trade deficit of US on the products so as to reduce the possible antidumping.

29

References

Blonigen, Bruce and Bown, Chad. Antidumping and Retaliation Threats [J]. Journal of International Economics, 2003, 60: 249-273.

Grubel, Herbert and Lloyd, Peter (1975), Intra-Industry Trade: the Theory and the Measurement of International Trade in Differentiated Products, London: Macmillan.

Shen Guobing (2007),“Revealed Comparative Advantages, Intra-Industry Trade, and Sino-U.S. Bilateral Trade Balance”(in Chinese), Management World 2, pp.5-16.

Shen Guobing (2008),“Antidumping and U.S.-China Bilateral Intra-Industry Trade: An Empirical Analysis”(in Chinese), World Economy Study 3, pp.48-55.

Stevenson, Cliff. Global Trade Protection Report 2007: Data & Analysis [R]. 2007: 1-20, www.antidumpingpublishing.com.

USITC (2006), “Import Injury Investigations Historical Case Statistics (FY 1980-2005)”, pp.13-24, http://www.usitc.gov.

30

Thanks a lot!