Bilateral Fiber and Textile Trade Documentation

This documentation for the bilateral fiber and textile trade database contains information on the source and availability of global trade data, product categories, and countries and regions. Also included is a discussion of inconsistencies in reported imports and exports and the methodology for reconciling bilateral trade flows.

Note: This data product is no longer being updated.

Source and Availability of Global Trade Data

The source of data used in the bilateral fiber and textile trade database is the United Nations (U.N.) Commodity Trade Statistics Database (Comtrade) assembled by the U.N. Statistics Division (UNSD). Comtrade compiles commodity trade statistics from national custom areas reporting to the United Nations. It contains the longest and most complete bilateral time-series trade data in terms of country and commodity coverage available from any other source. (UNSD uses "commodity" to describe trade of goods, both agricultural and nonagricultural. The bilateral fiber and textile trade database uses "product" to describe aggregates of textile and clothing goods.) The United Nations converts country-reported data expressed in local currency to nominal U.S. dollars.

Country availability of data depends on when data are reported by countries to UNSD and in what classification. Comtrade currently contains three commodity classifications for the international Harmonized System (HS) of trade codes, including HS-1988/92, HS-1996, and HS-2002. Prior to 1988, data were classified based on the Standard International Trade Classification (SITC), which was used by all countries in the U.N. system. The bilateral fiber and textile trade database is based on the HS-1988/92 classification because it provides the most detail for the longest period of time.

Comtrade has all reported data in their original classification and data converted from the original HS classification to other HS classifications. For example, South Africa reported its trade in 2002 using the HS-2002 classification. UNSD converted the data to both the HS-1996 classification and the HS-1988/92 classification. Thus, South African trade data are available for the full time-series starting in 1992. However, not all countries started reporting on the HS system at the same time. Russia, for example, did not start reporting trade on the HS system until 1996. Vietnam still reports trade using the SITC system and has yet to implement the HS classification.

There were 224 trading partners in Comtrade for 2002. They included both official countries and statistical trading territories. Of these trading partners, there were 132 reporting partners using the HS classification and 92 nonreporting partners.

Textiles and clothing are frequently shipped from one producing country to another country for redistribution to final destination countries. This is generally referred to as "re-export" activity. At the global level, re-exports can result in the double-counting of traded goods. In this database, exports do not include re-exports from other countries and imports are "retained imports" for home use.

There are two types of reporting systems in Comtrade: the "general trade system" and the "special trade system." Under the U.N.'s general system of reporting, re-exports are reported separately from national exports. Hong Kong, for example, reports its trade under the general system where the level of re-exports to it partners is reported separately from its bilateral national exports. Under the special trade system, countries do not report re-exports or imports destined for re-exports. For example, Singapore and the Netherlands report under the special system where their re-exports are not reported in Comtrade. Although the level of re-exports is not known from Comtrade, their reported exports are mainly reported as national exports.

Product Categories in the Database

The bilateral fiber and textile trade database is focused on the fiber-based content of traded textile-related products. Various textile and clothing aggregates differentiate among products produced from cotton, other natural fibers, manmade fibers, and natural-synthetic blends. These aggregates, created from Comtrade data using HS trade codes at the 6-digit level, include textiles and clothing, clothing, textiles, yarn, fabric for clothing, home furnishings, and industrial products. The database also contains information about trade in fibers (cotton, other natural fibers, and synthetics) and in capital inputs used to produce textile and clothing products (textile yarn machinery, weaving and knitting equipment, and auxiliary textile machinery). In addition, a merchandise trade total is included, enabling analysts to compare trade in textiles and/or clothing with overall trade. In total, data for 43 product aggregates are included in the database, specifically total merchandise trade and 42 product categories.

Product categories in the database
Total Merchandise
  Textile and Clothing: All
  Textile and Clothing: Cotton
  Textile and Clothing: Other Natural
  Textile and Clothing: Synthetic
  Textile and Clothing: Blended
    Clothing: All
      Clothing: Cotton
      Clothing: Other Natural
      Clothing: Synthetic
      Clothing: Blended
    Textiles: All
      Textiles: Cotton
      Textiles: Other Natural
      Textiles: Synthetic
      Textiles: Blended
      Yarn: All
        Yarn: Cotton
        Yarn: Other Natural
        Yarn: Synthetic
      Fabric: All
        Fabric: Cotton
        Fabric: Other Natural
        Fabric: Synthetic
        Fabric: Blended
      Home Furnishings: All
        Home Furnishings: Cotton
        Home Furnishings: Other Natural
        Home Furnishings: Synthetic
        Home Furnishings: Blended
      Industrial Products: All
        Industrial Products: Cotton
        Industrial Products: Other Natural
        Industrial Products: Synthetic
        Industrial Products: Blended
  Fiber: All
    Fiber: Cotton
    Fiber: Other Natural
    Fiber: Synthetic
  Machinery and Equipment
    Machinery: Yarn
    Equipment: Weaving and Knitting
    Machinery: Other Textile

The product categories are derived from 870 6-digit HS codes. The following tables Excel icon (16x16) identify the product content of each of the 42 categories by HS code and product description:

Product Coverage

Aggregate trade statistics from the bilateral fiber and textile trade database are not the same as those used by other organizations, even though Comtrade is a common data source. One reason is that this database makes a clear distinction between traded inputs, namely fibers and machinery, used in the production of intermediate and final textile and clothing products. Typically, manmade fibers (but not natural fibers, such as cotton, silk, or wool) are included in textile and/or clothing aggregates.

Another reason the bilateral fiber and textile trade database differs is the all-inclusive nature of its product coverage. This database includes all fibers, both natural and synthetic, as well as all textile and clothing products. Other studies, including publications issued by the World Trade Organization (WTO), the Central Intelligence Agency (CIA), and the Global Trade Analysis Project (GTAP), have less comprehensive product coverage for their textile and clothing sectors. The WTO report, The Global Textile and Clothing Industry Post the Agreement on Textiles and Clothing, Discussion Papers No. 5 (ISBN 92-870-1244-x), uses the WTO composition of textile and apparel trade, which consists of 795 HS 6-digit codes. In a March 2004 report published by the CIA, Textile and Apparel Exports: A Trade Profile (OTI FP 2004-001), 811 HS 6-digit codes were used to define trade in the sector. The GTAP database defines textiles and apparel using International Standard Industrial Classification (ISIC) codes. ISIC codes are linked to 801 HS codes via a production-trade concordance.

Countries and Regions in the Database

This database contains bilateral trade data for the principal regions and countries that export and import textiles and clothing. The 42 country/regional classifications in the database provide users with data on total world trade; individual countries and country groupings that add up to regional and world totals; and other country groupings, such as importers with Multifiber Arrangement (MFA) quotas.

World totals. Two world totals are identified in the database, one including and the other excluding intra-European Union (EU) trade. Although total world trade, including intra-EU trade, is widely used in the popular press, analysts often prefer to exclude trade between EU member countries because of their common border policies. For this reason, the trade share calculations for industrialized countries and importers with MFA-quotas aggregates do not include intra-EU trade.

Country and regional aggregates. Trade data are available for an industrialized country and a developing country aggregate. In the industrialized country aggregate, a complete picture of partner trade is available for three countries (the United States, Canada, and Japan) and two regions (the EU-15 and all other industrialized countries). Within the developing countries aggregate, a comprehensive set of partner data are available for five geographical areas:

More detailed information about partner trade exists for country/regions in some of the five geographical areas. For Africa and the Middle East, a comprehensive set of partner trade data are available for Turkey and three regions: North Africa and all other Middle East, African Growth and Opportunity Act (AGOA) eligible countries, and all other Sub-Saharan Africa. The AGOA was signed into U.S. law on May 18, 2000. Under this act, "AGOA-eligible" countries are allowed to sell to the U.S. market duty-free textile products made of fabrics manufactured in the United States  from U.S. yarn. Under a special provision, "lesser developed beneficiary countries" (those with a per capita gross national product (GNP) under $1,500 in 1998) have been allowed duty-free access for apparel made from fabric originating anywhere in the world. This special provision, which was set to expire on September 30, 2004, was granted a 3-year extension with the signing of the AGOA Acceleration Act on July 13, 2004.

For Asia, a comprehensive set of partner data exist for China, Hong Kong, Macau, South Korea, Taiwan, and three additional geographical breakdowns: South Asia, Southeast Asia, and all other North Asia. Within the South Asia and Southeast Asia areas, a complete set of partner data are available for Bangladesh, India, Pakistan, Indonesia, Thailand, and an all-other-South-Asia aggregate, as well as an all-other-Southeast-Asia aggregate.

For Latin America, a complete set of partner data exist for Mexico and three country groupings: Caribbean Basin Trade Partnership eligible countries, Andean Trade Preference Act countries, and an all-other-Latin-America-and-Caribbean aggregate. The U.S. Caribbean Basin Trade Partnership Act of 2000 provides duty- and quota-free treatment for apparel made in the beneficiary countries from fabrics manufactured in the United States from U.S. yarn. The U.S.-Andean Trade Preference Act of 2002 also exempts member countries from U.S. duties and import quotas for designated textile and apparel articles, provided that they are made from fabrics manufactured in the United States from U.S. yarn.

The system of regional aggregation used in this study contains no overlaps. Thus, the data for any country/region can be summed to arrive at regional and world totals. For example, the data for Bangladesh, India, Pakistan, and all other South Asia sum to that for South Asia. Data for China, Hong Kong, Macau, South Korea, Taiwan, South Asia, Southeast Asia, and all other North Asia sum to the total for Asia. Data for Africa and the Middle East, Asia, Eastern Europe, Former Soviet Union, and Latin America sum to the developing countries total. Data for developing countries and industrialized countries sum to the totals for the world, including and excluding intra-EU trade.

Regions and countries in the database Countries within the region (not in the database separately)
Industrialized countries  
  United States  
  European Union-15 Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, Netherlands, Portugal, Spain, Sweden, United Kingdom
  All other industrialized countries Australia, Bermuda, Faroe Islands, Gibraltar, Greenland, Iceland, New Zealand, Norway, Switzerland
Developing countries  
  Africa and the Middle East  
    North Africa and all other Middle East Algeria, Bahrain, Egypt, Iran, Iraq, Israel, Jordan, Kuwait, Lebanon, Libya, Morocco, Oman, Qatar, Saudi Arabia, Syria, Tunisia, United Arab Emirates, Yemen
    African Growth and Opportunity Act eligible Benin, Botswana, Cameroon, Cape Verde, Ethiopia, Ghana, Ivory Coast, Kenya, Lesotho, Madagascar, Malawi, Mali, Mauritius, Mozambique, Namibia, Niger, Nigeria, Rwanda, Senegal, Sierre Leone, South Africa, Swaziland, Tanzania, Uganda, Zambia
    All other Sub-Saharan Africa Angola, British Indian Ocean Territory, Burkina Faso, Burundi, Central African Republic, Chad, Comoros, Congo, Djibouti, Equatorial Guinea, Eritrea, Gabon, Gambia, Guinea, Guinea-Bissau, Liberia, Mauritania, Mayotte, Saint Helena, Sao Tome and Principe, Seychelles, Somalia, Sudan, Togo, Western Sahara, Zaire
    China (mainland)  
    Hong Kong  
    South Korea  
    South Asia  
      All other South Asia Bhutan, Maldives, Myanmar, Nepal, Sri Lanka
    Southeast Asia  
      All other Southeast Asia Brunei Darussalam, Cambodia, Christmas Island, Cocos (Keeling) Islands, Cook Islands, East Timor, Fiji, French Polynesia, Kiribati, Laos, Malaysia, Marshall Islands, Micronesia, Nauru, New Caledonia, Niue, Norfolk Island, Northern Mariana Islands, other U.S. territories in the Pacific, Palau, Papua New Guinea, Philippines, Pitcairn Islands, Samoa, Singapore, Solomon Islands, Tokelau, Tonga, Tuvalu, Vanuatu, Vietnam, Wallis and Futuna Islands
    All other North Asia Mongolia, North Korea
  Eastern Europe Albania, Bosnia and Herzegovina, Bulgaria, Croatia, Cyprus, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Macedonia, Malta, Poland, Romania, Slovakia, Slovenia, Yugoslavia
  Former Soviet Union Armenia, Azerbaijan, Belarus, Georgia, Kazakhstan, Kyrgyzstan, Moldova, Russia, Tajikistan, Turkmenistan, Ukraine, Uzbekistan
  Latin America  
    Caribbean Basin Trade Partnership  eligible Antigua, Aruba, Bahamas, Barbados, Belize, British Virgin Islands, Costa Rica, Dominica, Dominican Republic, El Salvador, Grenada, Guatemala, Haiti, Honduras, Jamaica, Montserrat, Netherlands Antilles, Nicaragua, St. Kitts and Nevis, St. Lucia, St. Vincent and  the Grenadines, Trinidad and Tobago
    Andean Trade Preference Act Bolivia, Colombia, Ecuador, Peru
    All other Latin America and Caribbean Anguilla, Argentina, Brazil, Cayman Islands, Chile, Cuba, Falkland Islands, Guyana, Panama, Paraguay, Saint Pierre and Miquelon, Suriname, Turks and Caicos Islands, Uruguay, Venezuela

Other areas. In addition to the comprehensive, nonoverlapping set of country/regional breakdowns in the table above, the database also contains data for some countries and country groupings of interest to users, such as the EU accession States.  "Other areas" in the database includes:

  • Central America Free Trade Agreement (CAFTA): Central American countries that are members of CAFTA;
  • Asian newly industrialized countries (NICs);
  • EU accession States: 15 countries from  Eastern Europe and the Former Soviet Union that joined the EU in 2004 or are scheduled to join in the future;
  • Importers with MFA quotas: importers that had quota agreements in 2004 with individual exporting countries for specific textile and clothing items; and
  • Non-WTO exporters of textiles and apparel: countries that export textiles and apparel, but are not members of the World Trade Organization (WTO).
"Other areas" in the database Countries within the area
Central America Free Trade Agreement Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua
Asian NICs (newly industrialized countries) Hong Kong, Macau, Singapore, South Korea, Taiwan
EU accession States Bosnia and Herzegovina, Croatia, Cyprus, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Macedonia, Malta, Poland, Slovakia, Slovenia, Yugoslavia
Importers with MFA quotas Austria, Belgium, Canada, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, Netherlands, Norway, Portugal, Spain, Sweden, United Kingdom, United States
Non-WTO T&A (textile and apparel) exporters Belarus, Cambodia, Laos, Nepal, Russia, Ukraine, Vietnam

Inconsistencies in Reported Imports and Exports

Discrepancies between reported exports and reported imports exist at the global level and for individual bilateral trade flows. Differences in trade flows are commonly found with all trading partners in both industrialized and developing regions. Reasons for these discrepancies include misidentified partners, misclassification of goods, under- or over-invoicing of goods, and differences in the valuation of goods by individual countries.

In 2002, reported imports of textiles and clothing were $372 billion, exceeding reported exports by $41 billion. Importers reported a higher level of global trade than exporters, except for within the European Union (EU), where reported exports were higher than reported imports. For trade outside the EU, exporters reported a lower value than what importers reported.

Reported exports of
textiles and clothing in 2002
Reported imports of
textiles and clothing in 2002
$ billion
World (excluding intra-EU trade)
Intra-EU trade
Source: Comtrade.

Reconciling Bilateral Trade Flows

To construct an analytical database of bilateral trade flows (such as this database), a country's total imports should be the same as its partners' total exports. When this condition is met for all countries, global imports equal global exports. Trade shares can be generated consistently from the database once these conditions are satisfied.

For each bilateral trade flow in this database, either the reported imports or the reported exports is used to represent the actual trade flow between two trading partners. Two basic cases present themselves: 1) one country reports but its partner fails to report, and 2) both countries report but each reports a different value for the same transaction. In the first case, we use the reported value, either from the importer or the exporter.

In the second case, we construct reliability indexes to reconcile differences in bilateral trade flows. For each transaction containing a discrepancy, it is assumed that there was a reporting error made by one of the reporting countries.  Reporting errors are detected and quantified statistically. Usually, the existence of a single error indicates similar errors made with other trading partners. The more frequent the error occurs, the less reliable the reporter becomes. For example, when a reporter misidentifies partners or misclassifies goods, they are deemed "less reliable" than a reporter that accurately reports a greater share of its trade flows with its partners. For a given transaction, reconciled data is used from the reporter deemed "more reliable."

Reliability index:
Reliability of importer (RIM) = imports accurately reported with partner / total reported imports
Reliability of exporter (RIX) = exports accurately reported with partner / total reported exports

Conditions for reconciling:
if RIM > RIX then reconciled trade value is the reported import value
if RIM < RIX then reconciled trade value is the reported export value

This procedure is implemented at the most disaggregated commodity level possible for all reporting importers and exporters. Calculations are made for all HS 6-digit commodities to generate reliability indices for a country as an exporter and as an importer. In all instances, the more reliable reporter is chosen for each bilateral trade flow. For the product aggregates in this database, a given trade flow will comprise trade that was reported by both the importer and the exporter. For example, clothing and textiles from Taiwan to Japan is reported by Taiwan as an exporter for some HS commodities but also reported by Japan as an importer for other HS commodities. This is because Taiwan is deemed to be the more reliable reporter for some commodities, while Japan is considered to be a more reliable reporter for others.

Using "reconciled" data means that the trade flows found in the bilateral fiber and textile trade database cannot be viewed as "imports" or "exports" but as reported trade flows deemed most reliable. The only case where a given trade flow might be reported solely by the exporter or the importer is when either the importer or the exporter fails to report or when the exporter or importer is deemed a more reliable reporter for all transactions.

The value of trade for product and regional totals in the bilateral fiber and textile trade database tend to be larger than those reported by other sources, in part because partner data is substituted for unreported trade at the most disaggregated (HS-6 digit) level. This database is unique in that exports equal imports for any given product aggregate and country/region.