Currently, both China and the United States calculate their external trade according to the principle of origin of goods. A growing number of international economists and statisticians believe, however, that there are irrationalities in this currently widely-used international statistical principle, especially because big errors might occur when it is applied to calculating entrepot trade and processing trade. Here lies the essence of the matter when the United States seriously exaggerates its trade deficit against China and distorts bilateral trade balance.

Statistics by origin refers to the determination of the source of the imports as the place where they were grown, manufactured or pro"icessed to make substantial changes. As a tool for countries to exercise trade management, the rules of origin have been important in the history of world trade. Today it is still widely applied to executing multinational trade accords and implementing foreign trade policy measures by different countries. On how to judge whether goods concerned have undergone ``substantial changes,'' however, there lacks a unified and detailed criterion for its application. The ``International Convention on Customs Procedures of Simplification and Harmonization'' stipulated by the Customs Co-operation Council in 1973 contains in its appendix one annex on the rules of origin, which indicates only principles but no enforceable provisions. The Uruguay Round negotiations of the General Agreement on Tariffs and Trade (GATT) reached the ``Agreement on Rules of Origin,'' aiming at co-ordinating member countries' rules of origin on non-preferential imports. The job to formulate the technical standards on the co-ordination principle was entrusted to the World Customs Organization but has not been accomplished till today. Due to the lack of internationally unified rules of origin, countries formulate rules of origin according to their own needs, resulting in different criteria, leaving room for discretion.

It deserves to be specially noted that the limitations of the old rules of origin are being increasingly exposed by the development of the world economy and great changes in the global economic pattern. In the past when cross-border trade and investment were scarce and commodity exchange ties among countries relatively simple, statistics based on the origin of goods could grossly reflect the division of labour, trade relations and corresponding pattern of interests among countries. Today, propelled by fast-growing economic and trade co-operation among countries in the world and cross-border investment that is increasing by the day, goods exchanged through international trade are no longer products made by one country, but ``world products'' whose manufacture involves efforts from several countries. It is obviously difficult for the current rules of origin, as employed by trade statistics, to accurately reflect the main changes in world economic development; these rules could even result in a distorted picture of trade balance situation.

Processing trade that has boomed in some countries and regions over the last two to three decades had the problem further complicated. Processing trade means a country importing main raw materials and auxiliary parts to be processed or assembled and re-exported. According to current rules, the country is deemed the place of origin because within it the imported goods have undergone substantial changes. Due to the fact that main raw materials and auxiliary parts are imported, however, the processing country often profits little from the trade. This was well illustrated in the ``Barbie doll'' example cited in the ``Barbie and the World Economy'' story which was run on the September 22, 1996 edition of the Los Angles Times. The story said that in the United States, the retail price of a ``Barbie'' toy imported from China was 9.99 US dollars compared to an import price of 2 US dollars. Of the 2 US dollars, China gained 35 US cents in service fee, another 65 US cents were spent on importing raw materials and still another 1 US dollar covers costs of transportation and management. To account the 2 US dollars as the revenue of Chinese exports to the United States according to rules of origin is clearly unreasonable.

Processing trade expansion is a major factor behind China's export growth in the 1990s. Total export volume rose to 151.1 billion US dollars in 1996 from 62.1 billion US dollars in 1990, up 16% per year on average. In the period, processing trade soared from 25.42 billion US dollars to 84.33 billion US dollars, an annual rise of 22.1%. In the 1990-96 period, the ratio of processing trade to overall Chinese exports rose from 41% to 55.8% and even amounted to 70% in its share of Chinese exports to the United States in 1996. The bulk of the sector has developed since the mid-1980s when investors from developed countries including the United States and Japan as well as Singapore, the Republic of Korea, and the Hong Kong and Taiwan regions started to move their labour-intensive industries or production procedures to China in a bid to cut production cost and enhance international competitiveness. Dependent on overseas investors' original marketing channels, processing trade products are primarily sold to their traditional markets including the United States via entrepot trade through Hong Kong. To calculate them according to certain rules of origin, China -- instead of the investors' home countries or regions and exporters of materials and auxiliary parts -- became the exporter. According to statistics published by the US Department of Commerce and reports from the US-China Business Council, the United States' trade deficit against Singapore, the Republic of Korea and the Hong Kong and Taiwan regions fell from 34 billion US dollars to 7.8 billion US dollars in the 1987-95 period, and in the meantime its trade deficit against China rocketed from 2.8 billion US dollars to 33.8 billion US dollars. These figures reflected the aforementioned process of place-of-origin transfer and showed that the trade balance between the United States and the countries and regions in Asia as a whole has not substantially changed over the years. In recent years, more than two-thirds of US-bound Chinese exports have undergone entrepot trade in the Hong Kong region. The value added at Hong Kong has been far greater than that in China's mainland, hence Hong Kong has benefited far more than the mainland from the processing trade. If that part of the added value is counted as Hong Kong exports, China's trade surplus against the United States will drop accordingly, or even turn into deficit.

Hence it is not difficult to conclude that,

1) While applying the origin-based statistics, it is imperative to be fully aware of the method's limitations and discern from the figures the actual benefits of all trading parties after careful analysis. That is the only way conducive to removing misunderstanding and properly settling bilateral trade disputes.

2) It is also a must to note the trend of increasingly closer international economic links and ever-growing cross-border investment and service trade that are growing by the day, and on that ground improve and perfect the computing of trade figures so that statistics will comply with reality and better serve the mutually beneficial co-operation among countries in the world.