Domestic Value Added in Chinese Exports

Hiau Looi Kee and Heiwai Tang

World Bank and Tufts University

December 2011

The �ndings and results of this paper do not represent the views of theWorld Bank, its Executive Directors, or the countries they represent

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 1 / 32

Designed by Apple in California, Assembled in China

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 2 / 32

US-China Trade, 1995 and 2010

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 3 / 32

Introduction

The widening bilateral imbalance in US-China trade has attracted alot of heated discussions, mainly focus on the US labor market impactof Chinese exports

Autor, Dorn and Hanson (2011) �nd that Chinese imports into the USsigni�cantly lower job creation, wages, and labor market participation

Scott (2011): �growing US trade de�cit with China costs 2.8 millionjobs between 2001 and 2010.�

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 4 / 32

Introduction

The widening bilateral imbalance in US-China trade has attracted alot of heated discussions, mainly focus on the US labor market impactof Chinese exports

Autor, Dorn and Hanson (2011) �nd that Chinese imports into the USsigni�cantly lower job creation, wages, and labor market participation

Scott (2011): �growing US trade de�cit with China costs 2.8 millionjobs between 2001 and 2010.�

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 4 / 32

Introduction

The widening bilateral imbalance in US-China trade has attracted alot of heated discussions, mainly focus on the US labor market impactof Chinese exports

Autor, Dorn and Hanson (2011) �nd that Chinese imports into the USsigni�cantly lower job creation, wages, and labor market participation

Scott (2011): �growing US trade de�cit with China costs 2.8 millionjobs between 2001 and 2010.�

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 4 / 32

Factory of the World

China being the �factory of the World,�due to its participation in theglobal production chain, particularly at the �nal assembly stage

While many products are �made in China�, they embody inputs fromaround the world

For Apple�s iPod, only US$4 out of its value of US$150 can beattributable to producers located in China, with the rest being createdmostly in the US, Japan, and Korea (Dedrick, Kraemer and Linden,2009)

Majority of Chinese exports are under the processing trade regime,whereby �rms in China import materials or parts to be furtherassembled, processed and exported

Domestic value added in Chinese exports may be far less than actualgross exports

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 5 / 32

Factory of the World

China being the �factory of the World,�due to its participation in theglobal production chain, particularly at the �nal assembly stage

While many products are �made in China�, they embody inputs fromaround the world

For Apple�s iPod, only US$4 out of its value of US$150 can beattributable to producers located in China, with the rest being createdmostly in the US, Japan, and Korea (Dedrick, Kraemer and Linden,2009)

Majority of Chinese exports are under the processing trade regime,whereby �rms in China import materials or parts to be furtherassembled, processed and exported

Domestic value added in Chinese exports may be far less than actualgross exports

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 5 / 32

Factory of the World

China being the �factory of the World,�due to its participation in theglobal production chain, particularly at the �nal assembly stage

While many products are �made in China�, they embody inputs fromaround the world

For Apple�s iPod, only US$4 out of its value of US$150 can beattributable to producers located in China, with the rest being createdmostly in the US, Japan, and Korea (Dedrick, Kraemer and Linden,2009)

Majority of Chinese exports are under the processing trade regime,whereby �rms in China import materials or parts to be furtherassembled, processed and exported

Domestic value added in Chinese exports may be far less than actualgross exports

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 5 / 32

Factory of the World

China being the �factory of the World,�due to its participation in theglobal production chain, particularly at the �nal assembly stage

While many products are �made in China�, they embody inputs fromaround the world

For Apple�s iPod, only US$4 out of its value of US$150 can beattributable to producers located in China, with the rest being createdmostly in the US, Japan, and Korea (Dedrick, Kraemer and Linden,2009)

Majority of Chinese exports are under the processing trade regime,whereby �rms in China import materials or parts to be furtherassembled, processed and exported

Domestic value added in Chinese exports may be far less than actualgross exports

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 5 / 32

Factory of the World

China being the �factory of the World,�due to its participation in theglobal production chain, particularly at the �nal assembly stage

While many products are �made in China�, they embody inputs fromaround the world

For Apple�s iPod, only US$4 out of its value of US$150 can beattributable to producers located in China, with the rest being createdmostly in the US, Japan, and Korea (Dedrick, Kraemer and Linden,2009)

Majority of Chinese exports are under the processing trade regime,whereby �rms in China import materials or parts to be furtherassembled, processed and exported

Domestic value added in Chinese exports may be far less than actualgross exports

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 5 / 32

0.553 0.5520.553

0.552

0.554

0.548

0.527

020

040

060

080

010

00Bi

llion

 USD

2000 2002 2004 2006year

Total Processing Exports Processing ShareRegular Exports Total Exports

China's Aggregate Exports by Type

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 6 / 32

What this paper is all about?

This paper provides evidence on the trend and the pattern of thedomestic value added in Chinese exports

Existing research includes Chen, Chang, Fung and Lau (2001),Hummels, Ishii and Yi (2001), Koopman, Wang and Wei (2008), andKoopman, Powers, Wang and Wei (2010), Johnson and Noguera(forthcoming)

These studies mainly rely on input-output tables at the sector level toassess foreign content in Chinese exports

This paper takes a ground-up approach: we use transaction-leveltrade data and �rm-level production data to assess the domestic valueadded in Chinese exports from 2000 to 2006

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 7 / 32

What this paper is all about?

This paper provides evidence on the trend and the pattern of thedomestic value added in Chinese exports

Existing research includes Chen, Chang, Fung and Lau (2001),Hummels, Ishii and Yi (2001), Koopman, Wang and Wei (2008), andKoopman, Powers, Wang and Wei (2010), Johnson and Noguera(forthcoming)

These studies mainly rely on input-output tables at the sector level toassess foreign content in Chinese exports

This paper takes a ground-up approach: we use transaction-leveltrade data and �rm-level production data to assess the domestic valueadded in Chinese exports from 2000 to 2006

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 7 / 32

What this paper is all about?

This paper provides evidence on the trend and the pattern of thedomestic value added in Chinese exports

Existing research includes Chen, Chang, Fung and Lau (2001),Hummels, Ishii and Yi (2001), Koopman, Wang and Wei (2008), andKoopman, Powers, Wang and Wei (2010), Johnson and Noguera(forthcoming)

These studies mainly rely on input-output tables at the sector level toassess foreign content in Chinese exports

This paper takes a ground-up approach: we use transaction-leveltrade data and �rm-level production data to assess the domestic valueadded in Chinese exports from 2000 to 2006

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 7 / 32

What this paper is all about?

This paper provides evidence on the trend and the pattern of thedomestic value added in Chinese exports

Existing research includes Chen, Chang, Fung and Lau (2001),Hummels, Ishii and Yi (2001), Koopman, Wang and Wei (2008), andKoopman, Powers, Wang and Wei (2010), Johnson and Noguera(forthcoming)

These studies mainly rely on input-output tables at the sector level toassess foreign content in Chinese exports

This paper takes a ground-up approach: we use transaction-leveltrade data and �rm-level production data to assess the domestic valueadded in Chinese exports from 2000 to 2006

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 7 / 32

How is it done?

1 Identify a set of �rms that only participated in processing exports in asingle HS 2-digit industry: all imports are used to produce processingexports within the industry

2 Merge these �rms to their production data from China�s NBS3 Compare import to export ratio of customs data to material to salesratio of production data ) weed out �rms that specialized inimporting or exporting

4 Use the clean sample to compute a �rm�s domestic value added ratio(DVAR) of the year, which simply equals the ratio of net exports togross exports at the �rm level

5 Aggregate up all the �rms in our clean sample to the HS 2 industriesthey are operating to calculate the industry DVAR based on the �rmsexports and imports data

6 Use export weighted average of the industry DVARs to �gure out thedomestic value added of �rms operating across multiple industries

7 Aggregate up all �rms to calculate domestic value added by exportdestinations

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 8 / 32

How is it done?

1 Identify a set of �rms that only participated in processing exports in asingle HS 2-digit industry: all imports are used to produce processingexports within the industry

2 Merge these �rms to their production data from China�s NBS

3 Compare import to export ratio of customs data to material to salesratio of production data ) weed out �rms that specialized inimporting or exporting

4 Use the clean sample to compute a �rm�s domestic value added ratio(DVAR) of the year, which simply equals the ratio of net exports togross exports at the �rm level

5 Aggregate up all the �rms in our clean sample to the HS 2 industriesthey are operating to calculate the industry DVAR based on the �rmsexports and imports data

6 Use export weighted average of the industry DVARs to �gure out thedomestic value added of �rms operating across multiple industries

7 Aggregate up all �rms to calculate domestic value added by exportdestinations

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 8 / 32

How is it done?

1 Identify a set of �rms that only participated in processing exports in asingle HS 2-digit industry: all imports are used to produce processingexports within the industry

2 Merge these �rms to their production data from China�s NBS3 Compare import to export ratio of customs data to material to salesratio of production data ) weed out �rms that specialized inimporting or exporting

4 Use the clean sample to compute a �rm�s domestic value added ratio(DVAR) of the year, which simply equals the ratio of net exports togross exports at the �rm level

5 Aggregate up all the �rms in our clean sample to the HS 2 industriesthey are operating to calculate the industry DVAR based on the �rmsexports and imports data

6 Use export weighted average of the industry DVARs to �gure out thedomestic value added of �rms operating across multiple industries

7 Aggregate up all �rms to calculate domestic value added by exportdestinations

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 8 / 32

How is it done?

1 Identify a set of �rms that only participated in processing exports in asingle HS 2-digit industry: all imports are used to produce processingexports within the industry

2 Merge these �rms to their production data from China�s NBS3 Compare import to export ratio of customs data to material to salesratio of production data ) weed out �rms that specialized inimporting or exporting

4 Use the clean sample to compute a �rm�s domestic value added ratio(DVAR) of the year, which simply equals the ratio of net exports togross exports at the �rm level

5 Aggregate up all the �rms in our clean sample to the HS 2 industriesthey are operating to calculate the industry DVAR based on the �rmsexports and imports data

6 Use export weighted average of the industry DVARs to �gure out thedomestic value added of �rms operating across multiple industries

7 Aggregate up all �rms to calculate domestic value added by exportdestinations

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 8 / 32

How is it done?

1 Identify a set of �rms that only participated in processing exports in asingle HS 2-digit industry: all imports are used to produce processingexports within the industry

2 Merge these �rms to their production data from China�s NBS3 Compare import to export ratio of customs data to material to salesratio of production data ) weed out �rms that specialized inimporting or exporting

4 Use the clean sample to compute a �rm�s domestic value added ratio(DVAR) of the year, which simply equals the ratio of net exports togross exports at the �rm level

5 Aggregate up all the �rms in our clean sample to the HS 2 industriesthey are operating to calculate the industry DVAR based on the �rmsexports and imports data

6 Use export weighted average of the industry DVARs to �gure out thedomestic value added of �rms operating across multiple industries

7 Aggregate up all �rms to calculate domestic value added by exportdestinations

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 8 / 32

How is it done?

1 Identify a set of �rms that only participated in processing exports in asingle HS 2-digit industry: all imports are used to produce processingexports within the industry

2 Merge these �rms to their production data from China�s NBS3 Compare import to export ratio of customs data to material to salesratio of production data ) weed out �rms that specialized inimporting or exporting

4 Use the clean sample to compute a �rm�s domestic value added ratio(DVAR) of the year, which simply equals the ratio of net exports togross exports at the �rm level

5 Aggregate up all the �rms in our clean sample to the HS 2 industriesthey are operating to calculate the industry DVAR based on the �rmsexports and imports data

6 Use export weighted average of the industry DVARs to �gure out thedomestic value added of �rms operating across multiple industries

7 Aggregate up all �rms to calculate domestic value added by exportdestinations

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 8 / 32

How is it done?

1 Identify a set of �rms that only participated in processing exports in asingle HS 2-digit industry: all imports are used to produce processingexports within the industry

2 Merge these �rms to their production data from China�s NBS3 Compare import to export ratio of customs data to material to salesratio of production data ) weed out �rms that specialized inimporting or exporting

4 Use the clean sample to compute a �rm�s domestic value added ratio(DVAR) of the year, which simply equals the ratio of net exports togross exports at the �rm level

5 Aggregate up all the �rms in our clean sample to the HS 2 industriesthey are operating to calculate the industry DVAR based on the �rmsexports and imports data

6 Use export weighted average of the industry DVARs to �gure out thedomestic value added of �rms operating across multiple industries

7 Aggregate up all �rms to calculate domestic value added by exportdestinations

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 8 / 32

Results

Domestic value added in Chinese processing exports was around 50percent in 2000, and it gradually increases to 62 percent in 2006

Such increase is wide spread across HS-2 industries as well as acrossdestination countries

Firm level regressions con�rm that export processing �rms substituteimported materials with domestic materials which explains the risingDVAR

Rising DVAR is not driven by rising domestic production costs

This result suggests that China may be moving up the globalproduction chain and is no longer only responsible for the �nal stageof productions

However, foreign content remains high in Chinese exports ) policyanalysis based on gross trade �ows will grossly overestimate theimpact of Chinese exports

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 9 / 32

Results

Domestic value added in Chinese processing exports was around 50percent in 2000, and it gradually increases to 62 percent in 2006

Such increase is wide spread across HS-2 industries as well as acrossdestination countries

Firm level regressions con�rm that export processing �rms substituteimported materials with domestic materials which explains the risingDVAR

Rising DVAR is not driven by rising domestic production costs

This result suggests that China may be moving up the globalproduction chain and is no longer only responsible for the �nal stageof productions

However, foreign content remains high in Chinese exports ) policyanalysis based on gross trade �ows will grossly overestimate theimpact of Chinese exports

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 9 / 32

Results

Domestic value added in Chinese processing exports was around 50percent in 2000, and it gradually increases to 62 percent in 2006

Such increase is wide spread across HS-2 industries as well as acrossdestination countries

Firm level regressions con�rm that export processing �rms substituteimported materials with domestic materials which explains the risingDVAR

Rising DVAR is not driven by rising domestic production costs

This result suggests that China may be moving up the globalproduction chain and is no longer only responsible for the �nal stageof productions

However, foreign content remains high in Chinese exports ) policyanalysis based on gross trade �ows will grossly overestimate theimpact of Chinese exports

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 9 / 32

Results

Domestic value added in Chinese processing exports was around 50percent in 2000, and it gradually increases to 62 percent in 2006

Such increase is wide spread across HS-2 industries as well as acrossdestination countries

Firm level regressions con�rm that export processing �rms substituteimported materials with domestic materials which explains the risingDVAR

Rising DVAR is not driven by rising domestic production costs

This result suggests that China may be moving up the globalproduction chain and is no longer only responsible for the �nal stageof productions

However, foreign content remains high in Chinese exports ) policyanalysis based on gross trade �ows will grossly overestimate theimpact of Chinese exports

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 9 / 32

Results

Domestic value added in Chinese processing exports was around 50percent in 2000, and it gradually increases to 62 percent in 2006

Such increase is wide spread across HS-2 industries as well as acrossdestination countries

Firm level regressions con�rm that export processing �rms substituteimported materials with domestic materials which explains the risingDVAR

Rising DVAR is not driven by rising domestic production costs

This result suggests that China may be moving up the globalproduction chain and is no longer only responsible for the �nal stageof productions

However, foreign content remains high in Chinese exports ) policyanalysis based on gross trade �ows will grossly overestimate theimpact of Chinese exports

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 9 / 32

Results

Domestic value added in Chinese processing exports was around 50percent in 2000, and it gradually increases to 62 percent in 2006

Such increase is wide spread across HS-2 industries as well as acrossdestination countries

Firm level regressions con�rm that export processing �rms substituteimported materials with domestic materials which explains the risingDVAR

Rising DVAR is not driven by rising domestic production costs

This result suggests that China may be moving up the globalproduction chain and is no longer only responsible for the �nal stageof productions

However, foreign content remains high in Chinese exports ) policyanalysis based on gross trade �ows will grossly overestimate theimpact of Chinese exports

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 9 / 32

Road Map Ahead

Data

Methodology1 Multiproduct �rms2 Negative domestic value added

Results

Conclusions

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 10 / 32

Data

The main data set covers the universe of Chinese import and exporttransactions in each month between 2000 and 2006

It reports values (in US dollars) of imports and exports at the HS8-digit level (over 7000 products) from a �rm to/from each country(over 200 destination and source countries)

We also use �rm-level data from the annual surveys of industrial �rmsconducted by China�s National Bureau of Statistics (NBS) for the1998-2007 period

The surveys cover all state-owned �rms and all non-state-owned �rmswith sales above 5 million yuan

The data set contains detailed information for about 100 variables,including �rm ID, address, ownership, output, value added, four-digitindustry code (480 categories), six-digit geographic code, exports,employment, original value of �xed asset, and intermediate inputs

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 11 / 32

Data

The main data set covers the universe of Chinese import and exporttransactions in each month between 2000 and 2006

It reports values (in US dollars) of imports and exports at the HS8-digit level (over 7000 products) from a �rm to/from each country(over 200 destination and source countries)

We also use �rm-level data from the annual surveys of industrial �rmsconducted by China�s National Bureau of Statistics (NBS) for the1998-2007 period

The surveys cover all state-owned �rms and all non-state-owned �rmswith sales above 5 million yuan

The data set contains detailed information for about 100 variables,including �rm ID, address, ownership, output, value added, four-digitindustry code (480 categories), six-digit geographic code, exports,employment, original value of �xed asset, and intermediate inputs

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 11 / 32

Data

The main data set covers the universe of Chinese import and exporttransactions in each month between 2000 and 2006

It reports values (in US dollars) of imports and exports at the HS8-digit level (over 7000 products) from a �rm to/from each country(over 200 destination and source countries)

We also use �rm-level data from the annual surveys of industrial �rmsconducted by China�s National Bureau of Statistics (NBS) for the1998-2007 period

The surveys cover all state-owned �rms and all non-state-owned �rmswith sales above 5 million yuan

The data set contains detailed information for about 100 variables,including �rm ID, address, ownership, output, value added, four-digitindustry code (480 categories), six-digit geographic code, exports,employment, original value of �xed asset, and intermediate inputs

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 11 / 32

Data

The main data set covers the universe of Chinese import and exporttransactions in each month between 2000 and 2006

It reports values (in US dollars) of imports and exports at the HS8-digit level (over 7000 products) from a �rm to/from each country(over 200 destination and source countries)

We also use �rm-level data from the annual surveys of industrial �rmsconducted by China�s National Bureau of Statistics (NBS) for the1998-2007 period

The surveys cover all state-owned �rms and all non-state-owned �rmswith sales above 5 million yuan

The data set contains detailed information for about 100 variables,including �rm ID, address, ownership, output, value added, four-digitindustry code (480 categories), six-digit geographic code, exports,employment, original value of �xed asset, and intermediate inputs

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 11 / 32

Data

The main data set covers the universe of Chinese import and exporttransactions in each month between 2000 and 2006

It reports values (in US dollars) of imports and exports at the HS8-digit level (over 7000 products) from a �rm to/from each country(over 200 destination and source countries)

We also use �rm-level data from the annual surveys of industrial �rmsconducted by China�s National Bureau of Statistics (NBS) for the1998-2007 period

The surveys cover all state-owned �rms and all non-state-owned �rmswith sales above 5 million yuan

The data set contains detailed information for about 100 variables,including �rm ID, address, ownership, output, value added, four-digitindustry code (480 categories), six-digit geographic code, exports,employment, original value of �xed asset, and intermediate inputs

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 11 / 32

Figure 1: Chinese exports by type, 2000-2006

0.553 0.5520.553

0.552

0.554

0.548

0.527

020

040

060

080

010

00Bi

llion

 USD

2000 2002 2004 2006year

Total Processing Exports Processing ShareRegular Exports Total Exports

China's Aggregate Exports by Type

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 12 / 32

Table 1: Top 10 Destinations for Chinese ProcessingExports

2000 2003 2006

Rank Share $ (Bil) Share $ (Bil) Share $ (Bil)

1 United States 0.25 25 United States 0.25 50.1 United States 0.25 113

2 Hong Kong 0.22 22.2 Hong Kong 0.23 44.6 Hong Kong 0.23 105

3 Japan 0.18 18.2 Japan 0.15 30.1 Japan 0.1 47.1

4 Korea 0.04 4.17 Germany 0.04 8.65 Germany 0.05 21.3

5 Germany 0.04 3.8 Netherlands 0.04 7.85 Netherlands 0.04 18.9

6 Singapore 0.03 3.17 Korea 0.04 7.81 Korea 0.04 18.2

7 Netherlands 0.03 3.07 Singapore 0.02 4.92 Singapore 0.03 12.8

8 UK 0.03 2.77 UK 0.02 4.81 UK 0.02 11.1

9 Taiwan 0.02 2.06 Taiwan 0.02 4.45 Taiwan 0.02 9.89

10 France 0.02 1.59 France 0.02 3.94 Malaysia 0.02 7.17

Total 101 198 449

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 13 / 32

Figure 2: Share of processing exports in top 10 destinations

0.2

.4.6

.8

US HK JP KR DE NL GB SG TW IT

Share of Processing Exports by Top 10 Destinations

2000 2006

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 14 / 32

Figure 3: Shares of processing exports in HS 2 industries

0.2

.4.6

.81

0.2

.4.6

.81

0.2

.4.6

.81

0.2

.4.6

.81

2 3 4 5 6 7 8 9 101112131415 16 17 18 19 20 21 22 23 24 25 26 27

2829303132333435363738 39 40 41 42 43 44 45 46 47 48 49

5051525354555657585960616263 64 65 66 67 68 69 70 71 7273747576787980818283

84 85 86 87 88 89 90 91 92 93 94 95 96 97

01­05: Animal & Animal Products 06­15: Vegetable Products 16­24: Foodstuffs 25­27: Mineral Products

28­38: Chemicals & Allied Industries 39­40: Plastics/ Rubbers41­43: Raw Hides, Skins, Leather, Furs44­49: Wood & Wood Products

50­63: Textiles 64­67: Footwear / Headgear 68­71: Stone / Glass 72­83: Metals

84­85: Machinery / Electrical 86­89: Transportation 90­97: Miscellaneous

Graphs by group_desc

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 15 / 32

In Summary

Processing trade is wide spread across industries and exportdestinations

For many destinations, such as the US, processing trade is the mainform of exports from China

For many industries, such as Electronics, Transportation, PlasticProducts, Toys, processing trade is the main form of exports fromChina

Given the high imported content in processing exports, any analysisbased on gross trade �ows can be highly misleading

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 16 / 32

In Summary

Processing trade is wide spread across industries and exportdestinations

For many destinations, such as the US, processing trade is the mainform of exports from China

For many industries, such as Electronics, Transportation, PlasticProducts, Toys, processing trade is the main form of exports fromChina

Given the high imported content in processing exports, any analysisbased on gross trade �ows can be highly misleading

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 16 / 32

In Summary

Processing trade is wide spread across industries and exportdestinations

For many destinations, such as the US, processing trade is the mainform of exports from China

For many industries, such as Electronics, Transportation, PlasticProducts, Toys, processing trade is the main form of exports fromChina

Given the high imported content in processing exports, any analysisbased on gross trade �ows can be highly misleading

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 16 / 32

In Summary

Processing trade is wide spread across industries and exportdestinations

For many destinations, such as the US, processing trade is the mainform of exports from China

For many industries, such as Electronics, Transportation, PlasticProducts, Toys, processing trade is the main form of exports fromChina

Given the high imported content in processing exports, any analysisbased on gross trade �ows can be highly misleading

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 16 / 32

Methodology

Start with a textbook identity, where the total revenue of a �rm(PY ) , consists of the following components: pro�ts, (π) , wages(wL) , cost of capital (rK ) , cost of domestic material

�PDMD

�and

cost of imported material�P IM I

�PY = π + wL+ rK + PDMD + P IM I

In theory, for processing trade �rms,

PY = EXP

P IM I = IMP

Hence domestic value added (DVA) for processing trade �rms is

EXP � IMP = π + wL+ rK + PDMD = DVA

Domestic value added ratio (DVAR) is therefore

DVAR =DVAEXP

=EXP � IMP

EXP= 1� IMP

EXP

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 17 / 32

Methodology

Start with a textbook identity, where the total revenue of a �rm(PY ) , consists of the following components: pro�ts, (π) , wages(wL) , cost of capital (rK ) , cost of domestic material

�PDMD

�and

cost of imported material�P IM I

�PY = π + wL+ rK + PDMD + P IM I

In theory, for processing trade �rms,

PY = EXP

P IM I = IMP

Hence domestic value added (DVA) for processing trade �rms is

EXP � IMP = π + wL+ rK + PDMD = DVA

Domestic value added ratio (DVAR) is therefore

DVAR =DVAEXP

=EXP � IMP

EXP= 1� IMP

EXP

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 17 / 32

Methodology

Start with a textbook identity, where the total revenue of a �rm(PY ) , consists of the following components: pro�ts, (π) , wages(wL) , cost of capital (rK ) , cost of domestic material

�PDMD

�and

cost of imported material�P IM I

�PY = π + wL+ rK + PDMD + P IM I

In theory, for processing trade �rms,

PY = EXP

P IM I = IMP

Hence domestic value added (DVA) for processing trade �rms is

EXP � IMP = π + wL+ rK + PDMD = DVA

Domestic value added ratio (DVAR) is therefore

DVAR =DVAEXP

=EXP � IMP

EXP= 1� IMP

EXP

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 17 / 32

Methodology

Start with a textbook identity, where the total revenue of a �rm(PY ) , consists of the following components: pro�ts, (π) , wages(wL) , cost of capital (rK ) , cost of domestic material

�PDMD

�and

cost of imported material�P IM I

�PY = π + wL+ rK + PDMD + P IM I

In theory, for processing trade �rms,

PY = EXP

P IM I = IMP

Hence domestic value added (DVA) for processing trade �rms is

EXP � IMP = π + wL+ rK + PDMD = DVA

Domestic value added ratio (DVAR) is therefore

DVAR =DVAEXP

=EXP � IMP

EXP= 1� IMP

EXP

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 17 / 32

Two Important Assumptions

1 we assume that there is no imported content in domestic materials,such that PDMD embodies purely domestic contents

2 we assume that imported materials have no Chinese contents, suchthat IMP is completely foreign made

If violated, the �rst assumption will lead to an over-estimation of DVA

If violated, the second assumption will lead to an under-estimation ofDVA

The overall bias due to the two assumptions is not clear, but there isnothing we can do at this stage to assess the direction of bias.

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 18 / 32

Two Important Assumptions

1 we assume that there is no imported content in domestic materials,such that PDMD embodies purely domestic contents

2 we assume that imported materials have no Chinese contents, suchthat IMP is completely foreign made

If violated, the �rst assumption will lead to an over-estimation of DVA

If violated, the second assumption will lead to an under-estimation ofDVA

The overall bias due to the two assumptions is not clear, but there isnothing we can do at this stage to assess the direction of bias.

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 18 / 32

Two Important Assumptions

1 we assume that there is no imported content in domestic materials,such that PDMD embodies purely domestic contents

2 we assume that imported materials have no Chinese contents, suchthat IMP is completely foreign made

If violated, the �rst assumption will lead to an over-estimation of DVA

If violated, the second assumption will lead to an under-estimation ofDVA

The overall bias due to the two assumptions is not clear, but there isnothing we can do at this stage to assess the direction of bias.

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 18 / 32

Two Important Assumptions

1 we assume that there is no imported content in domestic materials,such that PDMD embodies purely domestic contents

2 we assume that imported materials have no Chinese contents, suchthat IMP is completely foreign made

If violated, the �rst assumption will lead to an over-estimation of DVA

If violated, the second assumption will lead to an under-estimation ofDVA

The overall bias due to the two assumptions is not clear, but there isnothing we can do at this stage to assess the direction of bias.

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 18 / 32

Two Important Assumptions

1 we assume that there is no imported content in domestic materials,such that PDMD embodies purely domestic contents

2 we assume that imported materials have no Chinese contents, suchthat IMP is completely foreign made

If violated, the �rst assumption will lead to an over-estimation of DVA

If violated, the second assumption will lead to an under-estimation ofDVA

The overall bias due to the two assumptions is not clear, but there isnothing we can do at this stage to assess the direction of bias.

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 18 / 32

Multiproduct �rms

Above framework is helpful to �gure out DVAR at �rm level, but notuseful for product level or industry level analysis

Majority of �rms produce multiple products ) not be possible toassign the imported materials to each of the HS 6 digit products a�rm exports without knowing its �rm level production structure

Focus on the subset of export processing �rms that only operatewithin a single HS 2 industry

Example of HS 2 industries: Woven apparel products (HS 62),Non-woven apparel product (HS 61), Footwear, gaiters and the like(HS 64), Electrical machinery and equipment and parts (HS 85),Toys, games and sports requisites (HS 95)

For these set of single HS 2 industry processing �rms, all its importsare for exports within their HS 2 industry ) estimate the averageDVA for each HS 2 industry using the sample of single-HS2 exporters

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 19 / 32

Multiproduct �rms

Above framework is helpful to �gure out DVAR at �rm level, but notuseful for product level or industry level analysis

Majority of �rms produce multiple products ) not be possible toassign the imported materials to each of the HS 6 digit products a�rm exports without knowing its �rm level production structure

Focus on the subset of export processing �rms that only operatewithin a single HS 2 industry

Example of HS 2 industries: Woven apparel products (HS 62),Non-woven apparel product (HS 61), Footwear, gaiters and the like(HS 64), Electrical machinery and equipment and parts (HS 85),Toys, games and sports requisites (HS 95)

For these set of single HS 2 industry processing �rms, all its importsare for exports within their HS 2 industry ) estimate the averageDVA for each HS 2 industry using the sample of single-HS2 exporters

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 19 / 32

Multiproduct �rms

Above framework is helpful to �gure out DVAR at �rm level, but notuseful for product level or industry level analysis

Majority of �rms produce multiple products ) not be possible toassign the imported materials to each of the HS 6 digit products a�rm exports without knowing its �rm level production structure

Focus on the subset of export processing �rms that only operatewithin a single HS 2 industry

Example of HS 2 industries: Woven apparel products (HS 62),Non-woven apparel product (HS 61), Footwear, gaiters and the like(HS 64), Electrical machinery and equipment and parts (HS 85),Toys, games and sports requisites (HS 95)

For these set of single HS 2 industry processing �rms, all its importsare for exports within their HS 2 industry ) estimate the averageDVA for each HS 2 industry using the sample of single-HS2 exporters

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 19 / 32

Multiproduct �rms

Above framework is helpful to �gure out DVAR at �rm level, but notuseful for product level or industry level analysis

Majority of �rms produce multiple products ) not be possible toassign the imported materials to each of the HS 6 digit products a�rm exports without knowing its �rm level production structure

Focus on the subset of export processing �rms that only operatewithin a single HS 2 industry

Example of HS 2 industries: Woven apparel products (HS 62),Non-woven apparel product (HS 61), Footwear, gaiters and the like(HS 64), Electrical machinery and equipment and parts (HS 85),Toys, games and sports requisites (HS 95)

For these set of single HS 2 industry processing �rms, all its importsare for exports within their HS 2 industry ) estimate the averageDVA for each HS 2 industry using the sample of single-HS2 exporters

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 19 / 32

Multiproduct �rms

Above framework is helpful to �gure out DVAR at �rm level, but notuseful for product level or industry level analysis

Majority of �rms produce multiple products ) not be possible toassign the imported materials to each of the HS 6 digit products a�rm exports without knowing its �rm level production structure

Focus on the subset of export processing �rms that only operatewithin a single HS 2 industry

Example of HS 2 industries: Woven apparel products (HS 62),Non-woven apparel product (HS 61), Footwear, gaiters and the like(HS 64), Electrical machinery and equipment and parts (HS 85),Toys, games and sports requisites (HS 95)

For these set of single HS 2 industry processing �rms, all its importsare for exports within their HS 2 industry ) estimate the averageDVA for each HS 2 industry using the sample of single-HS2 exporters

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 19 / 32

Negative Domestic Value Added

About 10% of single-industry �rms have negative DVA, which is whenimports >> exports

Two problems that may cause processing imports to be greater thanprocessing exports within a �rm in a given year

1 Leakage issues2 Inventory management

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 20 / 32

Negative Domestic Value Added

About 10% of single-industry �rms have negative DVA, which is whenimports >> exports

Two problems that may cause processing imports to be greater thanprocessing exports within a �rm in a given year

1 Leakage issues2 Inventory management

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 20 / 32

Negative Domestic Value Added

About 10% of single-industry �rms have negative DVA, which is whenimports >> exports

Two problems that may cause processing imports to be greater thanprocessing exports within a �rm in a given year

1 Leakage issues

2 Inventory management

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 20 / 32

Negative Domestic Value Added

About 10% of single-industry �rms have negative DVA, which is whenimports >> exports

Two problems that may cause processing imports to be greater thanprocessing exports within a �rm in a given year

1 Leakage issues2 Inventory management

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 20 / 32

Leakage Issues

Some single-industry export processing �rms, while exporting, mayimport materials for other �rms

Hence, processing imports can be larger than processing exports at�rm level ) DVA < 0Merging the single-industry export processing �rms to theirproduction data collected by the census of manufacturing �rmsDrop �rms where there is excessive imports:

materialsrevenue

<IMPEXP

For the remaining �rms, we obtain the share of DVA in exports of theindustries they operate in by summing up the imports and exports atHS 2 industry levelFor those �rms that operate across multiple HS 2 industries, we applythe export weighted average DVA ratio of the industries they are in toobtain their �rm level DVA ratio

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 21 / 32

Leakage Issues

Some single-industry export processing �rms, while exporting, mayimport materials for other �rmsHence, processing imports can be larger than processing exports at�rm level ) DVA < 0

Merging the single-industry export processing �rms to theirproduction data collected by the census of manufacturing �rmsDrop �rms where there is excessive imports:

materialsrevenue

<IMPEXP

For the remaining �rms, we obtain the share of DVA in exports of theindustries they operate in by summing up the imports and exports atHS 2 industry levelFor those �rms that operate across multiple HS 2 industries, we applythe export weighted average DVA ratio of the industries they are in toobtain their �rm level DVA ratio

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 21 / 32

Leakage Issues

Some single-industry export processing �rms, while exporting, mayimport materials for other �rmsHence, processing imports can be larger than processing exports at�rm level ) DVA < 0Merging the single-industry export processing �rms to theirproduction data collected by the census of manufacturing �rms

Drop �rms where there is excessive imports:

materialsrevenue

<IMPEXP

For the remaining �rms, we obtain the share of DVA in exports of theindustries they operate in by summing up the imports and exports atHS 2 industry levelFor those �rms that operate across multiple HS 2 industries, we applythe export weighted average DVA ratio of the industries they are in toobtain their �rm level DVA ratio

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 21 / 32

Leakage Issues

Some single-industry export processing �rms, while exporting, mayimport materials for other �rmsHence, processing imports can be larger than processing exports at�rm level ) DVA < 0Merging the single-industry export processing �rms to theirproduction data collected by the census of manufacturing �rmsDrop �rms where there is excessive imports:

materialsrevenue

<IMPEXP

For the remaining �rms, we obtain the share of DVA in exports of theindustries they operate in by summing up the imports and exports atHS 2 industry levelFor those �rms that operate across multiple HS 2 industries, we applythe export weighted average DVA ratio of the industries they are in toobtain their �rm level DVA ratio

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 21 / 32

Leakage Issues

Some single-industry export processing �rms, while exporting, mayimport materials for other �rmsHence, processing imports can be larger than processing exports at�rm level ) DVA < 0Merging the single-industry export processing �rms to theirproduction data collected by the census of manufacturing �rmsDrop �rms where there is excessive imports:

materialsrevenue

<IMPEXP

For the remaining �rms, we obtain the share of DVA in exports of theindustries they operate in by summing up the imports and exports atHS 2 industry level

For those �rms that operate across multiple HS 2 industries, we applythe export weighted average DVA ratio of the industries they are in toobtain their �rm level DVA ratio

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 21 / 32

Leakage Issues

Some single-industry export processing �rms, while exporting, mayimport materials for other �rmsHence, processing imports can be larger than processing exports at�rm level ) DVA < 0Merging the single-industry export processing �rms to theirproduction data collected by the census of manufacturing �rmsDrop �rms where there is excessive imports:

materialsrevenue

<IMPEXP

For the remaining �rms, we obtain the share of DVA in exports of theindustries they operate in by summing up the imports and exports atHS 2 industry levelFor those �rms that operate across multiple HS 2 industries, we applythe export weighted average DVA ratio of the industries they are in toobtain their �rm level DVA ratio

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 21 / 32

Inventory Management

Current year processing imports may not be used fully for currentlyyear processing exports, which causes DVA < 0

For those industries that use a lot of commodities based materialssuch as iron, copper and crude oil, �rms have incentive to stock upimported materials when the international prices of such commoditiesare low in order to hedge again rising prices in the future

No easy solutions

But this may not be a big problem, as once we weed out �rms withexcessive imports, DVA are mostly positive

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 22 / 32

Inventory Management

Current year processing imports may not be used fully for currentlyyear processing exports, which causes DVA < 0

For those industries that use a lot of commodities based materialssuch as iron, copper and crude oil, �rms have incentive to stock upimported materials when the international prices of such commoditiesare low in order to hedge again rising prices in the future

No easy solutions

But this may not be a big problem, as once we weed out �rms withexcessive imports, DVA are mostly positive

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 22 / 32

Inventory Management

Current year processing imports may not be used fully for currentlyyear processing exports, which causes DVA < 0

For those industries that use a lot of commodities based materialssuch as iron, copper and crude oil, �rms have incentive to stock upimported materials when the international prices of such commoditiesare low in order to hedge again rising prices in the future

No easy solutions

But this may not be a big problem, as once we weed out �rms withexcessive imports, DVA are mostly positive

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 22 / 32

Inventory Management

Current year processing imports may not be used fully for currentlyyear processing exports, which causes DVA < 0

For those industries that use a lot of commodities based materialssuch as iron, copper and crude oil, �rms have incentive to stock upimported materials when the international prices of such commoditiesare low in order to hedge again rising prices in the future

No easy solutions

But this may not be a big problem, as once we weed out �rms withexcessive imports, DVA are mostly positive

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 22 / 32

Results

Domestic value added share in total processing export was 50% in2000

It has since risen to 62% in 2006

Rising DVAR is observed across industries and across destinations

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 23 / 32

Results

Domestic value added share in total processing export was 50% in2000

It has since risen to 62% in 2006

Rising DVAR is observed across industries and across destinations

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 23 / 32

Results

Domestic value added share in total processing export was 50% in2000

It has since risen to 62% in 2006

Rising DVAR is observed across industries and across destinations

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 23 / 32

Figure 5: Estimated average domestic value added ratio inChinese exports

.5.5

5.6

.65

2000 2002 2004 2006year

Arithmetic mean Export­weighted mean

Average domestic value added ratios of processing exports

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 24 / 32

Figure 6: Estimated domestic value added ratios by HS 2industries

0.2

.4.6

.81

0.2

.4.6

.81

0.2

.4.6

.81

0.2

.4.6

.81

3 7 8 16 19 20 21 22 28 29 30 32 33 34 38

39 40 41 42 43 44 48 49 5254555657585960616263

64 65 66 67 68 69 70 71 72 73 74 76 82 83 84 85

86 87 89 90 91 92 94 95 96

01­05: Animal & Animal Products 06­15: Vegetable Products 16­24: Foodstuffs 28­38: Chemicals & Allied Industries

39­40: Plastics/ Rubbers41­43: Raw Hides, Skins, Leather, Furs44­49: Wood & Wood Products 50­63: Textiles

64­67: Footwear / Headgear 68­71: Stone / Glass 72­83: Metals 84­85: Machinery / Electrical

86­89: Transportation 90­97: Miscellaneous

(Exp

­Imp)

/Exp

Graphs by group_desc

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 25 / 32

Figure 6: Distributions of domestic value added ratios

01

23

4kd

ensi

ty d

var

.2 .4 .6 .8 1x

2000 20032006

Distribution of domestic value added, all sectors

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 26 / 32

Figure 7: Domestic value added vs capital intensity

16:Preparations of19:Prep cereals, f

20:Prep of vegetab

21:Misc edible pre22:Beverages, spir

28:Inorganic chemi

29:Organic chemica

30:Pharmaceutical

32:Tanning or dyei

33:Essential oils

34:Soap, organic s

38:Miscellaneous c39:Plastics and ar

40:Rubber and arti

41:Raw hides and s

42:Articles of lea

43:Fur skins and a

44:Wood and articl

48:Paper & paperbo

49:Printed books,

52:Cotton

54:Man­made filame

55:Man­made staple

56:Wadding, felt a

57:Carpets and oth

58:Special woven f59:Impregnated, co

60:Knitted or croc

61:Apparel and clo

62:Apparel and clo63:Other made up t

64:Footwear, gaite65:Headgear and pa

66:Umbrellas, sun67:Prep feathers a

68:Articles of sto

69:Ceramic product

70:Glass and glass

71:Natural or cult72:Iron and steel

73:Articles of iro

74:Copper and arti

76:Aluminum and ar

82:Tools, implemen

83:Misc articles o84:Machinery and m

85:Electrical mach

86:Railway or tram

87:Vehicles other

89:Ships, boats an

90:Optical, photog

91:Clocks and watc

92:Musical instrum

94:Furniture; bedd

95:Toys, games and96:Misc manufactur

.2.4

.6.8

1

0 .5 1 1.5 2(mean) lcapint2

95% CI Fitted values(Exp­Imp)/Exp

slope = ­0.045; t stat =­1.62; N=56

DVA Ratios and Capital Intensity (2006)

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 27 / 32

Why is domestic value added ratio rising?

It could be due to rising production costs in China ) China is losingcompetitiveness

It could be because �rms substitution imported materials withdomestic materials ) China is moving up the value added chain inglobal production network

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 28 / 32

Why is domestic value added ratio rising?

It could be due to rising production costs in China ) China is losingcompetitiveness

It could be because �rms substitution imported materials withdomestic materials ) China is moving up the value added chain inglobal production network

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 28 / 32

Dependent variable: domestic value added ratio

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 29 / 32

Dependent variable: import to material ratio

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 30 / 32

In Summary

Regression results suggest that processing �rms are substitutingimported materials with domestic materials

Evidence of China move up the value added chain and no longer onlyresponsible for the �nal assembly of products

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 31 / 32

In Summary

Regression results suggest that processing �rms are substitutingimported materials with domestic materials

Evidence of China move up the value added chain and no longer onlyresponsible for the �nal assembly of products

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 31 / 32

Conclusions

We assess domestic value added in Chinese exports based on customstransaction data

Our results are similar to those existing papers that use aggregateinput-output table

Domestic value added only consists of about 60 percent of Chineseexports

Bilateral imbalance between US and China could be grosslyoverestimated based on gross trade �gures

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 32 / 32

Conclusions

We assess domestic value added in Chinese exports based on customstransaction data

Our results are similar to those existing papers that use aggregateinput-output table

Domestic value added only consists of about 60 percent of Chineseexports

Bilateral imbalance between US and China could be grosslyoverestimated based on gross trade �gures

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 32 / 32

Conclusions

We assess domestic value added in Chinese exports based on customstransaction data

Our results are similar to those existing papers that use aggregateinput-output table

Domestic value added only consists of about 60 percent of Chineseexports

Bilateral imbalance between US and China could be grosslyoverestimated based on gross trade �gures

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 32 / 32

Conclusions

We assess domestic value added in Chinese exports based on customstransaction data

Our results are similar to those existing papers that use aggregateinput-output table

Domestic value added only consists of about 60 percent of Chineseexports

Bilateral imbalance between US and China could be grosslyoverestimated based on gross trade �gures

Kee and Tang (WB and Tufts) Domestic Value Added in Chinese Exports 12/11 32 / 32