## Outline of the world input output tables of the WIOD database

## Source details

**Canonical URL:** [Outline of the world input output tables of the WIOD database](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16207.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16207.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16207.pdf.json)

---

### Introduction: context and motivation
- International trade over the last "20–25 years" characterized by an ever more flattening of the world (Antras, 2015).
- Key drivers:
  - Opening of China (accession to the WTO in "2001") and fall of the Iron Curtain increasing available low-wage workforce.
  - Liberal trade policies and falling transportation costs; Uruguay Round created the WTO in "1995".
  - Mean tariff rates decreased from "34 percent in 1996" to "2.9 percent in 2012" (World Development Indicators).
  - Number of regional trade agreements more than quadrupled from the mid 1990’s to today (WTO Secretariat).
  - Improvements in information and communication technologies (e.g., Internet) reducing coordination costs.
- Consequence: trade in intermediate goods now accounts for approximately "two thirds of world trade".
- Vertical specialization examples:
  - Apple’s iPod and Boeing’s Dreamliner 787: “More than 70 percent of the plane’s value is not generated by Boeing.”

### Policy relevance and institutional context
- Trade in value added is a central interest for international institutions and policy makers; a UN expert group (members from IMF, OECD, UN, World Bank, WTO) is working on a “Handbook on a System of Extended International Global Accounts.”
- IMF quota formula components and current openness measure:
  - GDP: "50 percent"
  - Openness: "30 percent" (calculated on gross exports and imports)
  - Economic variability: "15 percent"
  - International reserves: "5 percent"
- Improved value added data could provide a more accurate openness parameter for quota calculation by avoiding double counting, though currently available data do not allow robust adjustment of the openness quota.
- Understanding positions in global value chains (GVCs) matters for assessing vulnerability to shock transmission (example: Boehm et al., 2015 on the 2011 Tohoku earthquake).
- World Bank recommended three-step policy adjustment related to GVCs: attracting FDI, expanding/strengthening GVC participation, turning GVC participation into sustainable growth (Taglioni and Winkler, 2016).

### Data: WIOD coverage and characteristics
- WIOD used in the paper:
  - Time coverage: "1995 to 2011".
  - Industry coverage: "35 industries" (primary-, durable-, nondurable-, service- and finance industries).
  - Country coverage: "41 countries" (the 41st is the rest of the world modeled as one economy).
- WIOD includes the newly industrialized economies (NICs): Brazil, China, India, Indonesia, Mexico and Turkey — these countries account for more than "90 percent of GDP generated by newly industrialized countries."
- Construction and quality:
  - World input output tables compiled from national account statistics, supply and use tables, and international trade data in goods and services.
  - Data quality depends on national statistics; tradeoff between data coverage and data quality.
  - Comparative datasets: GTAP, TiVa, EORA, IDE-JETRO — IDE-JETRO has higher quality but limited country coverage; EORA covers many countries but lower comparative quality.
  - Needs: better national supply and use tables, more disaggregation (WIOD’s "35 industries" more aggregated than UN Comtrade), and standardized methods of measurement, compilation, and classification.
  - IMF role: provide assistance to countries with lower data quality.

### Structure of the WIOD input-output matrices (one-year outline)
- Intermediate use matrix:
  - Dimension: "1435*1435" (41 countries * 35 industries).
  - Rows: production (industry × origin country); Columns: intermediate use by industry × destination country.
  - Main diagonal 35*35 blocks: intermediate domestic consumption of domestic production.
  - Off-diagonal blocks: intermediate consumption of foreign production (intermediate exports/imports).
- Final use matrix:
  - Shows production used as final input by destination countries.
  - Final consumption contains five categories (e.g., capital formation, household consumption, government consumption).
  - Main diagonal blocks: domestic final consumption of domestically produced goods; off-diagonal: final consumption of foreign production.
- Total use and total supply must be identical.

### Computation: deriving value added trade measures
- Equilibrium total gross output starting point:
  - Equation [1]: y = A y + f, where y is total gross output, A is intermediate input coefficient matrix, f is total final demand.
- Intermediate coefficient matrix:
  - A obtained by element-wise division of intermediate input matrix A (absolute values) by total output vector TO (A ⊘ TO).
- Leontief inverse:
  - Rearranged equation [2]: y = (I − A)^{-1} f, with L = (I − A)^{-1}.
- Foreign value added embodied in exports:
  - Equation [3]: FVA_o = vac .* L * x_o
    - vac is the value added coefficient vector: VAC ./ TO, with zeros for industries in origin country o when computing FVA for o.
    - x_o is the export vector (sum of foreign intermediate and final consumption for the origin’s industries).
- Foreign consumption of domestic value added (FCDVA):
  - Equation [4]: FCDVA_o = vac .* (L * F_o)
    - F_o is the final demand matrix aggregated over final demand categories 1 to J for each destination country.
- Analytical focus:
  - Origin and share of foreign value added in exports; destination of foreign consumption of domestic value added.
  - Methodological references: Miller and Blair (2009); Foster-McGregor and Stehrer (2013); country applications include Stehrer and Stöllinger (2013) and Andersen et al. (2015).

### Limitations and data caveats
- World input output tables are compiled from national sources — cannot be measured directly; data quality varies by country and influences WIOD quality.
- European country bias in WIOD coverage reflects higher-quality national statistics in European countries.
- Tradeoff between coverage and quality: higher-quality datasets may have limited country coverage; broader datasets may reduce per-country reliability.
- Need to improve national supply and use tables and standardize compilation methods to enhance world input output tables.

### Key numeric findings and statistics (growth indices and shares)
- Table 1 (Growth of Total Exports and Imports, Value Added and Output 1995–2011; 1995=1)
  - Total Exports: Brazil 5.27; China 12.41; India 8.01; Indonesia 4.04; Mexico 4.26; Turkey 5.53
  - Total Imports: Brazil 4.7; China 12.6; India 8.61; Indonesia 3.71; Mexico 4.77; Turkey 6.43
  - Value Added: Brazil 3.11; China 10.06; India 5.08; Indonesia 3.51; Mexico 3.6; Turkey 3.28
  - Output: Brazil 3.17; China 11.79; India 4.96; Indonesia 3.59; Mexico 3.48; Turkey 3.71
- Table 2 (Ratio of Intermediate to Final Exports and Imports in 2011)
  - Intermediate-/Final Exports: Brazil 2.69; China 1.22; India 1.11; Indonesia 4.02; Mexico 1.63; Turkey 1.4
  - Intermediate-/Final Imports: Brazil 7.98; China 4.69; India 2.52; Indonesia 3.05; Mexico 1.97; Turkey 0.94
- China: domestic value added share in gross exports dropped from around 83 percent in 1995 to around 75 percent in 2011 (largest decrease ~8 percentage points); in 2011 25 percent of export value was generated abroad.
- 2008 financial crisis effect: on average domestic value added share in NICs increased by about 3 percent from 2008 to 2009.
- Mexico: domestic value added share <70 percent in 2011.
- Brazil: close to 90 percent domestic value added in gross exports in 2011.
- Foreign value added contributors to China (2011 shares of China’s foreign value added):
  - “Rest of the World”: >30 percent of China’s foreign value added share → around 7.5 percent of China’s total exports.
  - Japan: 10 percent; Korea: 6 percent; Taiwan: 6 percent.
- Mexico’s foreign value added origins:
  - 1995: >60 percent originated from the US.
  - 2011: US share around 37 percent; China share 17 percent (China ~1 percent in 1995).
- China’s contribution to value added exports increased by factors:
  - Brazil: increase of factor 7
  - Indonesia: increase of factor 4
  - India: increase of factor 5
  - Turkey: increase of factor 3
- Destination of foreign consumption of China’s value added (2011):
  - “Rest of the World” 22.5 percent; United States 22 percent; Japan 8.6 percent.
- Mexico 2011: almost 60 percent of foreign consumption of Mexico’s value added occurs in the US; Canada ~6 percent.
- Concentration and dispersion (Lorenz/Gini for thirty largest destination countries):
  - The thirty largest consumers cover more than 99 percent of total foreign consumption for all NICs.
  - Mexico’s foreign consumption of value added is the most concentrated (1995 and 2011).
  - 2011 dispersion ranking (highest dispersion first): Turkey, Brazil, India, China.
  - Turkey’s Gini decreased from 0.64 to 0.56 between 1995 and 2011 (Germany’s share fell from 27.5 percent to 11.6 percent; US share fell from 11.2 percent to 6.3 percent).

### Industry structure and measurement differences
- Gross output vs. value added discrepancies:
  - China 2011 example: Gross output indicates sector 30-33 (Electrical and Optical Equipment) accounts for 35 percent of China’s total export value; value added basis shows this sector accounts for around 12 percent of foreign Chinese value added consumption.
  - Electrical and optical equipment shows high shares of foreign value added — China often performs final assembly while components are produced elsewhere.
  - Primary and service industries show higher relative importance when trade is measured in value added rather than gross output.
- Value added export shares by industry category for six NICs (1995–2011) — selected shares (2011 and 1995 in parentheses where provided in Table A2):
  - Brazil: Primary 25.23 (15.93); Durables 7.94 (10.09); Nondurables 28.68 (38.68); Finance 4.62 (5.32); Service 33.52 (29.98)
  - China: Primary 12.79 (19.00); Durables 21.00 (14.16); Nondurables 34.88 (40.47); Finance 4.53 (4.50); Service 26.79 (21.87)
  - India: Primary 16.11 (22.51); Durables 14.11 (7.81); Nondurables 20.62 (34.77); Finance 7.36 (5.46); Service 41.81 (29.44)
  - Indonesia: Primary 46.85 (27.41); Durables 7.31 (7.19); Nondurables 27.34 (38.08); Finance 1.19 (7.24); Service 7.32 (20.08)
  - Mexico: Primary 27.23 (18.73); Durables 22.79 (20.26); Nondurables 17.84 (19.67); Finance 3.56 (5.57); Service 28.58 (35.77)
  - Turkey: Primary 10.94 (16.37); Durables 13.19 (10.42); Nondurables 35.98 (50.30); Finance 4.49 (4.94); Service 35.4 (17.97)
- Sectoral trends:
  - India: service and finance combined almost 50 percent in 2011.
  - China and Mexico: service + finance ~30 percent of total value added exports in 2011.
  - Turkey: services 35 percent and nondurables 36 percent in 2011; lowest primary share 11 percent.
  - Brazil, Mexico, Indonesia: highest value added share in primary products; Indonesia ~47 percent of total value added exports from primary products, mining ~37 percent.
  - For all NICs except Mexico, nondurables were the major contributor of value added exports in 1995; this share decreased for all NICs by an average of 25 percent, while shares of services (except Mexico) and durables increased.

### Summary findings and heterogeneity across NICs
- Traditional gross output trade measures have limitations in the presence of increased vertical specialization and intermediate trade; value added measures reveal large differences — exports shares measured by gross output vs. value added differed by more than 20 percent for some industries.
- All six NICs showed large increases in output, exports, imports and value added over 1995–2011.
- Origin and destination of foreign value added became more diverse from 1995 to 2011; NICs’ share of foreign value added in exports increased while the share of industrialized countries decreased.
- Country-specific positions:
  - Brazil and Indonesia: larger shares of domestic value added in gross exports (less integrated in global production).
  - China, India, Mexico, Turkey: increased participation in global production sharing.
  - India: major destination for service offshoring — almost 50 percent of India’s value added exports originate in services.
  - China: concentrates on manufacturing and assembly, with increased production of intermediates.
  - Mexico: production highly focused on the US.
  - Turkey: concentrates on the European market.

### Policy implications and recommendations
- Avoid “one size fits all” policies for NICs; tailor policy to each country’s role and position in global value chains.
- Suggested directions:
  - For countries not well integrated (e.g., Brazil): focus on attracting foreign capital to increase GVC participation.
  - For well-integrated countries (e.g., India, China): focus on sustainable growth and learning from GVC participation.
- Assessments of vulnerability and spillover risk should account for dispersion of foreign consumption of domestic value added:
  - Greater dispersion implies wider participation in trade — increases likelihood of shock transmission but spreads and may mitigate consequences compared to concentrated-export scenarios.
  - Demand from NICs and other emerging markets helped drive recovery after the 2008 financial crisis.

*Source: _wp16207 - 1. Outline of the world input output tables of the WIOD database; sections III and IV and Appendices as provided.*

### 1. Outline of the world input output tables of the WIOD database ........................................... 7

### Outline of the world input output tables of the WIOD database

### Introduction: context and motivation
- International trade over the last "20–25 years" characterized by an ever more flattening of the world (Antras, 2015).
- Key drivers cited:
  - Opening of China (accession to the WTO in "2001") and fall of the Iron Curtain increasing available low-wage workforce.
  - Liberal trade policies and falling transportation costs; Uruguay Round created the WTO in "1995".
  - Mean tariff rates decreased from "34 percent in 1996" to "2.9 percent in 2012" (World Development Indicators).
  - Number of regional trade agreements more than quadrupled from the mid 1990’s to today (WTO Secretariat).
  - Improvements in information and communication technologies (e.g., Internet) reducing coordination costs.
- Consequence: trade in intermediate goods now accounts for approximately "two thirds of world trade".
- Examples of vertical specialization:
  - Apple’s iPod (Dedrick et al., 2010) and Boeing’s Dreamliner 787 (Tang et al., 2009): a great proportion of component production is outsourced and a sizeable proportion of the product’s value added is generated abroad; “More than 70 percent of the plane’s value is not generated by Boeing.”
- Implications:
  - Gross trade measures fail to allocate value added along global value chains and cannot account for export value generated by imported intermediates.
  - Value added trade measures revise bilateral trade balances and comparative advantage (e.g., Johnson and Noguera (2012a)).

### Policy relevance and institutional context
- Trade in value added is of significant interest to international institutions and policy makers.
- UN expert group (members from IMF, OECD, UN, World Bank, WTO) working on a “Handbook on a System of Extended International Global Accounts.”
- For the IMF, improved value added data could provide a more accurate openness parameter for quota calculation by avoiding double counting, though currently available data do not allow robust adjustment of the openness quota.
  - Current quota formula components: GDP (50 percent), openness (30 percent), economic variability (15 percent), and international reserves (5 percent). Openness is calculated based on gross exports and imports.
- Understanding positions in global value chains is essential to assess vulnerability to shock transmission (example: Boehm et al., 2015 on the 2011 Tohoku earthquake).
- World Bank recommended three-step policy adjustment related to GVCs: attracting FDI, expanding/strengthening GVC participation, turning GVC participation into sustainable growth (Taglioni and Winkler, 2016).

### Data: WIOD coverage and characteristics
- This paper uses the World Input Output Database (WIOD) to compute value added indicators for newly industrialized economies (NICs).
- WIOD details:
  - Time coverage: "1995 to 2011".
  - Industry coverage: "35 industries" comprising primary-, durable-, nondurable-, service- and finance industries.
  - Country coverage: "41 countries" (the 41st is the rest of the world modeled as one economy).
  - WIOD includes the newly industrialized economies of Brazil, China, India, Indonesia, Mexico and Turkey — these countries account for more than "90 percent of GDP generated by newly industrialized countries."
- Construction and quality:
  - World input output tables are compiled from national statistics (national account statistics, supply and use tables) and international trade data in goods and services.
  - Data quality depends on national statistics; this creates a tradeoff between data coverage and data quality.
  - Examples of other datasets: GTAP, TiVa, EORA, IDE-JETRO. IDE-JETRO has higher quality but limited country coverage; EORA covers many countries but lower comparative quality.
  - Need for improvement: better national supply and use tables, more disaggregation (WIOD’s "35 industries" is more aggregated than UN Comtrade), and standardized methods of measurement, compilation, and classification.
  - IMF role: provide assistance to countries with lower data quality.

### Structure of the WIOD input-output matrices (one-year outline)
- Intermediate use matrix:
  - Dimension: "1435*1435" (41 countries * 35 industries).
  - Rows: production (industry × origin country); Columns: intermediate use by industry × destination country.
  - Main diagonal 35*35 blocks: intermediate domestic consumption of domestic production.
  - Off-diagonal blocks: intermediate consumption of foreign production (intermediate exports/imports highlighted in illustrative squares).
- Final use matrix:
  - Shows production used as final input by destination countries.
  - Final consumption contains five categories (e.g., capital formation, household consumption, government consumption).
  - Main diagonal blocks: domestic final consumption of domestically produced goods; off-diagonal: final consumption of foreign production.
- Total use and total supply must be identical.

### Computation: deriving value added trade measures
- Starting point (equilibrium total gross output):
  - Equation [1]: y = A y + f, where y is total gross output, A is intermediate input coefficient matrix, f is total final demand.
  - Interpretation: total gross output for industry n in origin country o equals intermediate demand A y plus final demand f.
- Construction of intermediate coefficient matrix:
  - A is obtained by element-wise division of intermediate input matrix A (absolute values) by total output vector TO (A ⊘ TO).
  - Example notation: A_o,n;o',n' elements and division by TO corresponding industry output.
- Leontief inverse:
  - Rearranged equation [2]: y = (I − A)^{-1} f, with L = (I − A)^{-1} being the Leontief inverse.
  - L shows how much production from each industry in each country is required given a vector of final good consumption.
- Foreign value added embodied in exports:
  - Equation [3]: FVA_o = vac .* L * x_o
    - vac is the value added coefficient vector: VAC ./ TO (value added per unit of total output) for all industries and countries, with zeros for industries in origin country o when computing FVA for o.
    - x_o is the export vector (sum of foreign intermediate and final consumption for the origin’s industries).
  - Interpretation: FVA_o is the product of value added coefficients, the Leontief inverse, and the export vector — yielding the foreign value added share for country o.
- Foreign consumption of domestic value added (FCDVA):
  - Equation [4]: FCDVA_o = vac .* (L * F_o)
    - F_o is the final demand matrix aggregated over final demand categories 1 to J for each destination country.
  - Interpretation: value added coefficients element-wise multiplied with the Leontief inverse times final demand yield the foreign consumption of domestic value added by destination.

### Analytical focus of the paper
- Main empirical focus: origin and share of foreign value added in exports; destination of foreign consumption of domestic value added.
- Methodological references and precedents:
  - Computations follow Miller and Blair (2009) and Foster-McGregor and Stehrer (2013).
  - Builds on country-level applications such as Stehrer and Stöllinger (2013) for Austria and Andersen et al. (2015) for Denmark.
- Application: comparative analysis of newly industrialized economies (Brazil, China, India, Indonesia, Mexico, Turkey) to assess differing roles in global production and global value chains despite similar development stages.

### Limitations and data caveats
- World input output tables cannot be measured directly; they are compiled from national sources — data quality varies by country and influences overall WIOD quality.
- European country bias in WIOD coverage partly reflects higher-quality national statistics in European countries.
- Tradeoff between coverage and quality: higher-quality datasets may have limited country coverage; broader datasets may reduce per-country reliability.
- Need to improve national supply and use tables and standardize compilation methods to enhance world input output tables.

*Source: _wp16207 - 1. Outline of the world input output tables of the WIOD database*

### 2. Equation [4] yields 퐹퐶퐷푉퐴

### 2. Equation [4] yields 퐹퐶퐷푉퐴

### III. THE POSITION OF NEWLY INDUSTRIALIZED ECONOMIES IN GLOBAL VALUE CHAINS

- When firms slice up their production, they determine which stage of production takes place in which country; upstream producers produce raw materials, downstream producers assemble finalized intermediate inputs into final products.
- Policy advice differs for upstream vs. downstream producers.

### A. Value Added, Trade and Output

- Table 1: Six NICs: Growth of Total Exports and Imports, Value Added and Output 1995–2011 (1995=1)
  - Total Exports: Brazil 5.27; China 12.41; India 8.01; Indonesia 4.04; Mexico 4.26; Turkey 5.53
  - Total Imports: Brazil 4.7; China 12.6; India 8.61; Indonesia 3.71; Mexico 4.77; Turkey 6.43
  - Value Added: Brazil 3.11; China 10.06; India 5.08; Indonesia 3.51; Mexico 3.6; Turkey 3.28
  - Output: Brazil 3.17; China 11.79; India 4.96; Indonesia 3.59; Mexico 3.48; Turkey 3.71
- China showed the largest increase for all four indicators over 1995–2011, followed by India.
- All six NICs reached pre-2008-crisis levels by 2010; some industrialized economies (e.g., Denmark) were still below pre-crisis levels in 2011.
- For all countries, value added and output grew less than exports and imports — interpreted as increased specialization.
- Table 2: Six NICs: Ratio of Intermediate to Final Exports in 2011
  - Intermediate- / Final Exports: Brazil 2.69; China 1.22; India 1.11; Indonesia 4.02; Mexico 1.63; Turkey 1.4
  - Intermediate- / Final Imports: Brazil 7.98; China 4.69; India 2.52; Indonesia 3.05; Mexico 1.97; Turkey 0.94
- In 2011 all countries exported more intermediate products than final products; India exported almost four times more intermediate than final products? (table shows Intermediate-/Final Exports 1.11 for India).
- Turkey imported more final than intermediate products (Intermediate-/Final Imports 0.94).
- China’s higher ratio of intermediate to final imports vs. exports suggests a downstream position (main assembler).

### B. Foreign and Domestic Value Added in Exports

- Using equation [3] yields the foreign value added content embodied in exports 퐹푉퐴
!
; Figure 2 shows domestic value added component (1-퐹푉퐴
!
) for six NICs and the World from 1995–2011.
- Domestic value added share decreased for all countries except Indonesia.
- China: domestic value added share in gross exports dropped from around 83 percent in 1995 to around 75 percent in 2011 (largest decrease ~8 percentage points); thus in 2011 25 percent of export value was generated abroad.
- The 2008 financial crisis interrupted the decreasing domestic value added trend: on average domestic value added share in NICs increased by about 3 percent from 2008 to 2009.
- Mexico: lowest domestic value added share of traded goods with less than 70 percent in 2011 — explained by Mexico’s role as an assembly location for US intermediate products (Maquiladoras).
- Brazil: close to 90 percent domestic value added in gross exports in 2011 — possibly due to a more protectionist tariff regime.
- All countries except Mexico show a larger share of domestic value added in gross exports than the world average.
- India’s domestic value added share advantage over world average fell from around 11 percentage points in 1995 to around 5 percentage points in 2011.
- Foreign value added shares in exports by partner country (altered equation 3):
  - China’s foreign value added share ~25 percent in 2011.
  - “Rest of the World” accounts for more than 30 percent of China’s foreign value added share → around 7.5 percent of China’s total exports.
  - Top foreign value added contributors to China include: Japan 10 percent; Korea 6 percent; Taiwan 6 percent.
- Mexico: foreign value added in exports highly concentrated in the US
  - 1995: >60 percent of foreign value added in Mexican exports originated from the US.
  - 2011: US share around 37 percent; China share 17 percent (China was ~1 percent in 1995).
- China’s contribution to value added exports increased for Brazil (increase of factor 7), Indonesia (increase of factor 4), India (increase of factor 5), and Turkey (increase of factor 3).
- The share of foreign value added from the other five NICs increased only slightly since 1995 and remained fairly low.

### C. Trade in Value Added - Foreign Consumption of Domestic Value Added (FCDVA)

- Tracking foreign consumption of domestic value added captures indirect trade linkages and interdependencies; equation 4 computes the foreign consumption of domestic value added (FCDVA).
- Figure 4: destination of foreign domestic value added consumption for China and Mexico, 1995 vs. 2011.
  - China 2011: “Rest of the World” 22.5 percent of total foreign consumption; United States 22 percent; Japan 8.6 percent.
  - Consumption destinations are largely driven by economic size and distance (gravity parameters).
- From 1995 to 2011 consumption of Chinese products became more dispersed; similar patterns for other NICs where consumption destinations became more dispersed and NICs/other emerging markets increased their consumption share while industrialized economies decreased theirs.
- Mexico 2011: almost 60 percent of foreign consumption of Mexico’s value added occurs in the US; Canada ~6 percent.
- For Mexico and India the major trading partner’s consumption share (the US) remained fairly stable from 1995 to 2011; India’s share even slightly increased.
- Figure 5: Lorenz Curve and Gini coefficient for concentration of the thirty largest destination countries of value added consumption:
  - The thirty largest consumers cover more than 99 percent of total foreign consumption for all NICs.
  - Mexico’s foreign consumption of value added is the most concentrated (both in 1995 and 2011).
  - 2011 dispersion ranking (highest dispersion first): Turkey, Brazil, India, China.
  - Turkey’s Gini decreased from 0.64 to 0.56 between 1995 and 2011 (Germany’s share fell from 27.5 percent to 11.6 percent; US share fell from 11.2 percent to 6.3 percent).
- Implications for spillover risk:
  - Greater dispersion implies more countries participate in trade — makes it unlikely that foreign shocks are not transmitted, but also spreads risks and may mitigate consequences compared to concentrated-export scenarios.
  - Example: demand from NICs and other emerging markets was a main driver for recovery after the 2008 financial crisis.
- Industry structure of value added exports:
  - Gross output vs. value added measurement differences are striking.
  - Figure 6 (China 2011): Gross output indicates sector 30-33 (Electrical and Optical Equipment) accounts for 35 percent of China’s total export value; value added basis shows this sector accounts for around 12 percent of foreign Chinese value added consumption.
  - Electrical and optical equipment shows high shares of foreign value added — China often performs final assembly while components are produced elsewhere.
  - Primary and service industries show higher relative importance when trade is measured in value added rather than gross output.
  - Evaluating exports purely on gross output can overestimate manufacturing relative importance and underestimate services and primary sectors.

- Figure 7: Value added export shares by industry category for six NICs (1995–2011)
  - Industry categories: primary (yellow), durable (light blue), nondurable (grey), finance (orange), service (dark blue).
  - India: large share of value added exports from service and finance; combined almost 50 percent in 2011.
  - China and Mexico: service + finance ~30 percent of total value added exports in 2011.
  - Turkey: services 35 percent and nondurables 36 percent; lowest primary share 11 percent.
  - Brazil, Mexico, Indonesia: highest value added share in primary products; Indonesia ~47 percent of total value added exports from primary products, mining ~37 percent.
  - For all NICs except Mexico, nondurables were the major contributor of value added exports in 1995; this share decreased for all NICs by an average of 25 percent, while shares of services (except Mexico) and durables increased.

### IV. SUMMARY AND CONCLUSION

- Traditional gross output trade measures have limitations given increased vertical specialization and dominance of intermediate trade; value added measures reveal large differences.
- Exports shares measured by gross output vs. value added differed by more than 20 percent for some industries.
- All NICs showed large increases in output, exports, imports and value added over 1995–2011.
- Origin and destination of foreign value added became more diverse from 1995 to 2011; NICs’ share of foreign value added in exports increased while the share of industrialized countries decreased.
- Heterogeneity across NICs:
  - Brazil and Indonesia show larger shares of domestic value added in gross exports (less integrated in global production).
  - China, India, Mexico, Turkey show increased participation in global production sharing.
  - India: major destination for service offshoring — almost 50 percent of India’s value added exports originate in services.
  - China: concentrates on manufacturing and assembly, but its production of intermediates increased substantially.
  - Mexico: production highly focused on the US.
  - Turkey: concentrates on the European market.
- Policy implication: avoid “one size fits all” policies for NICs; tailor policy to each country’s role and position in global value chains.
  - Example policy directions:
    - For countries not well integrated (e.g., Brazil): focus on attracting foreign capital to increase GVC participation.
    - For well-integrated countries (e.g., India, China): focus on sustainable growth and learning from GVC participation.

*Source: WIOD, Author’s calculations; content from the provided PDF excerpt.*

### 3.0 IGO.

### 3.0 IGO.

### References cited
- Tang, C.S., Zimmerman, J.D. and Nelson, J.I., 2009, January. Managing new product development and supply chain risks: The Boeing 787 case. In Supply Chain Forum: An International Journal (Vol. 10, No. 2, pp. 74-86). Taylor & Francis.
- UNECE, 2015, Draft Guide to Measuring Global Production, https://www.unece.org/fileadmin/DAM/stats/documents/ece/ces/bur/2015/February/16Add.1-Guide_to_Measuring_Global_Production.pdf
- World Trade Organization, 2001, Doha Ministerial Declaration. Ministerial Conference, Fourth Session. Washington DC: World Trade Organization, www.wto.org/english/thewto_e/minist_e/min01_e/mindecl_e.pdf.

### Appendix: figures and tables inventory
- Figure A1. 787 Dreamliner Suppliers; Source: Boeing, Reuters.
- Figure A2. Output, Exports, Imports and Value Added for China, India and Indonesia from 1995 to 2011. Source: WIOD, Author’s calculations.
- Figure A3. Output, Exports, Imports and Value Added for Mexico, Brazil and Turkey from 1995 to 2011. Source: WIOD, Author’s calculations.
- Figure A4. Foreign Value Added by Country in Percent of Total Foreign Value Added for Brazil, Indonesia, India and Turkey – 1995 and 2011. Source: WIOD, Author’s calculations.
- Figure A5. Foreign Consumption of Domestic Value Added for Brazil, Indonesia, India and Turkey by Partner Country – 1995 and 2011. Source: WIOD, Author’s calculations.
- Figure A6. Five NICs: Foreign Consumption of China’s Domestic Value Added Measured in Gross and Value Added Exports in 2011. Source: WIOD, Author’s calculations.
- Figure A7. Six NICs: Foreign Consumption of Domestic Value Shares by Industry – 1995 to 2011. Source: WIOD, Author’s calculations.
- Table A1. Six NICs Growth of Final and Intermediate Imports and Exports from 1995 to 2011 (1995=1), Base Value 1995 in Brackets (in Million USD). Source: WIOD, Author’s calculations.
- Table A2. Six NICs: Foreign Consumption of Domestic Value Shares by Industry Category - 1995 to 2011. Source: WIOD, Author’s calculations.
- Table A3. Industries and Industry Catergories. Source: Industry classification: NACE Rev. 1.

### Key numeric data from Table A1 (Six NICs: growth indices and base values, 1995=1; base values in brackets in Million USD)
- Brazil:
  - Final Exports: 3.21 (13,453)
  - Intermediate Exports: 5.05 (42,466)
  - Final Imports: 4.05 (24,864)
  - Intermediate Imports: 5.13 (38,761)
- China:
  - Final Exports: 10.33 (91,728)
  - Intermediate Exports: 14.84 (78,078)
  - Final Imports: 7.32 (43,027)
  - Intermediate Imports: 14.9 (99,119)
- India:
  - Final Exports: 9.19 (17,445)
  - Intermediate Exports: 7.18 (24,745)
  - Final Imports: 9.05 (11,826)
  - Intermediate Imports: 8.45 (31,916)
- Indonesia:
  - Final Exports: 2.75 (15,862)
  - Intermediate Exports: 4.58 (38,276)
  - Final Imports: 3.31 (14,602)
  - Intermediate Imports: 3.86 (38,246)
- Mexico:
  - Final Exports: 6.79 (31,769)
  - Intermediate Exports: 4.35 (48,989)
  - Final Imports: 6 (19,207)
  - Intermediate Imports: 4.32 (52,507)
- Turkey:
  - Final Exports: 3.98 (16,452)
  - Intermediate Exports: 7.67 (11,921)
  - Final Imports: 9.43 (12,738)
  - Intermediate Imports: 4.8 (23,570)

### Selected numeric details from Table A2 (Foreign Consumption of Domestic Value Shares by Industry Category - 1995 to 2011)
- Brazil:
  - Primary: 25.23 (15.93)
  - Durables: 7.94 (10.09)
  - Nondurables: 28.68 (38.68)
  - Finance: 4.62 (5.32)
  - Service: 33.52 (29.98)
- China:
  - Primary: 12.79 (19.00)
  - Durables: 21.00 (14.16)
  - Nondurables: 34.88 (40.47)
  - Finance: 4.53 (4.50)
  - Service: 26.79 (21.87)
- India:
  - Primary: 16.11 (22.51)
  - Durables: 14.11 (7.81)
  - Nondurables: 20.62 (34.77)
  - Finance: 7.36 (5.46)
  - Service: 41.81 (29.44)
- Indonesia:
  - Primary: 46.85 (27.41)
  - Durables: 7.31 (7.19)
  - Nondurables: 27.34 (38.08)
  - Finance: 1.19 (7.24)
  - Service: 7.32 (20.08)
- Mexico:
  - Primary: 27.23 (18.73)
  - Durables: 22.79 (20.26)
  - Nondurables: 17.84 (19.67)
  - Finance: 3.56 (5.57)
  - Service: 28.58 (35.77)
- Turkey:
  - Primary: 10.94 (16.37)
  - Durables: 13.19 (10.42)
  - Nondurables: 35.98 (50.30)
  - Finance: 4.49 (4.94)
  - Service: 35.4 (17.97)

### Industry classification (Table A3) — selected mappings (WIOD code, Industry Number, Description, Category)
- 1A, B — Agriculture, Hunting, Forestry and Fishing — primary
- 2C — Mining and Quarrying — primary
- 3 (15, 16) — Food, Beverages and Tobacco — nondurables
- 4 (17, 18) — Textiles and Textile Products — nondurables
- 5 (19) — Leather, Leather and Footwear — nondurables
- 6 (20) — Wood and Products of Wood and Cork — nondurables
- 7 (21, 22) — Pulp, Paper, Paper, Printing and Publishing — nondurables
- 8 (23) — Coke, Refined Petroleum and Nuclear Fuel — nondurables
- 9 (24) — Chemicals and Chemical Products — nondurables
- 10 (25) — Rubber and Plastics — nondurables
- 11 (26) — Other Non-Metallic Mineral — nondurables
- 12 (27, 28) — Basic Metals and Fabricated Metal — nondurables
- 13 (29) — Machinery, Nec — durables
- 14 (30-33) — Electrical and Optical Equipment — durables
- 15 (34, 35) — Transport Equipment — durables
- 16 (36, 37) — Manufacturing, Nec; Recycling — durables
- 17 (E) — Electricity, Gas and Water Supply — nondurables
- 18 (F) — Construction — service
- 19 (50) — Sale, Maintenance and Repair of Motor Vehicles and Motorcycles; Retail Sale of Fuel — service
- 20 (51) — Wholesale Trade and Commission Trade, Except of Motor Vehicles and Motorcycles — service
- 21 (52) — Retail Trade, Except of Motor Vehicles and Motorcycles; Repair of Household Goods — service
- 22 (H) — Hotels and Restaurants — service
- 23 (60) — Inland Transport — service
- 24 (61) — Water Transport — service
- 25 (62) — Air Transport — service
- 26 (63) — Other Supporting and Auxiliary Transport Activities; Activities of Travel Agencies — service
- 27 (64) — Post and Telecommunications — service
- 28 (J) — Financial Intermediation — finance
- 29 (70) — Real Estate Activities — service
- 30 (71-74) — Renting of M&Eq and Other Business Activities — service
- 31 (L) — Public Admin and Defence; Compulsory Social Security — service
- 32 (M) — Education — service
- 33 (N) — Health and Social Work — service
- 34 (O) — Other Community, Social and Personal Services — service
- 35 (P) — Private Households with Employed Persons — service

*Source: _wp16207 - 3.0 IGO.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16207.pdf_