شواهد از آمریکا و چین: چگونه فرهنگ تجارت خارجی را تحت تاثیر قرار می دهد

Using a gravity model of trade and the U.S. and Chinese panel data, I present evidence that supports the views that linguistic links have become more important in foreign trade than geographical proximity and that linguistic influence on trade is more significant in China than in the U.S. The estimated result based on a broad panel of economies shows little overall relation between religious similarity and foreign trade in both the U.S. and China. But there is an indication that the religious dissimilarity tends to retard foreign trade with poor countries and regions and to encourage foreign trade with richer places. I also found that, although religious retardation on foreign trade is more significant in China than in the U.S., religious dissimilarity tends to retard the US export more than that of China; by contrast, it tends to retard the Chinese import more than that of the U.S. At last, the socio-economic implications of the estimated results are discussed.

Since the 1980s, especially since the early 1990s when the Cold War came to an end, economic activities in the homogeneously cultural environment have become more and more important than in the heterogeneously cultural environments. For example, trade among the ASEAN, Taiwan, Hong Kong, South Korea, and the mainland China, most of which either fall within or are closely related to the Chinese cultural circle, increased from less than 10% to over 30% of their total trade from the 1950s to the 1990s. Similar shifts towards the intraregional trade also occurred in intra-Latin American trade in the early 1990s, with trade between Brazil and Argentina tripling and Colombia-Venezuela trade quadrupling between 1990 and 1993. In 1994, Brazil replaced the United States as Argentina's principal trading partner.1 No body has now doubted about the role of cultural factor in international trade. Quantitative studies, however, have not been conducted until the 1990s. Havrylyshyn and Pritchett (1991), for example, find that the three languages, Portuguese, Spanish, and English, are significant in decreasing order of magnitude. In the study of poor countries, Foroutan and Pritchett (1993) find that French, Spanish and English are statistically significant. After trying to supplement the general language term and allowing each of the major languages to have an independent extra coefficient, Frankel and Wei (1995) (Table 5) found that two languages, English and Chinese, appear to qualify as especially important.2 More recently, Frankel et al. (1997) and Rauch (1999) use nine languages (English, Spanish, Chinese, Arabic, French, German, Japanese, Dutch, and Portuguese)3 and treat international linguistic links as a dummy variable for whether countries of a pair speak a common language or had colonial links earlier in the twentieth century. Frankel et al.'s results show a highly significant effect when all the languages are constrained to have the same coefficient. The estimate fluctuates over time between 0.33 and 0.77. Pooled time-series estimates of the coefficient (in natural log) cluster around 0.44, which implies that two countries sharing linguistic/colonial links tend to trade roughly 55% (that is, exp (0.44) ≈ 1.55) more than they would otherwise (1997, pp. 74–75). There is a consistent interpretation that the above estimated coefficients on linguistic links exhibit a trend whereby trade in the post-war period took place among countries being linguistically similar to each other, or in other words, they interpret this as possible evidence of increased cultural barriers to trade. However, it should be noted that cultural variables have been highly simplified in the existing studies, probably due to the fact that the cultural factors were only treated as a complement variable in the determinants of trade. For example, the linguistic links between countries were only treated in the above studies as a dummy variable. As most countries are linguistically diversified, the international (or interregional) linguistic links should not be simply expressed by the numbers ‘1’ (for countries to share a common language) and ‘0’ (for otherwise countries).4 In addition, the above literature omits another cultural variable, religion, that could play, at least in some cases, a more important role in the determinants of economic development than the linguistic variable (Guo, 2001, Chapter 5; and Guo and Hwang, 2002). This paper is organized as follows. Section 2 develops a highly simplified framework in which international trade is determined jointly by economic complementarity and cultural similarity. Our interests focus on the clarification of various economic and cultural conditions under which trade may be either encouraged or retarded. The empirical analysis is based on a gravity model on trade, which includes two cultural variables: linguistic and religious. To generate the cross-national data of the cultural variables, I will use a comprehensive measure of cultural links between any pair of economies. Although the components of a culture have been variously defined, I will only focus on two cultural elements: language and religion. Of course, my discussion of these cultural elements is not definitive and perhaps would not satisfy anthropologists. Nevertheless, my consideration is due to the concerns that ‘language’ is an effective tool of communication and that ‘religion’ can provide the insights into the characteristics of a culture. Using the US and Chinese panel data and the Ordinary Least Squares (OLS) method, Section 3 estimates quantitatively the effects of cultural influences on foreign trade. The choice of the panel data is due to not only that the U.S. is the largest developed economy and China is the largest developing economy, but also that most cultural elements (language, religion, and others) of the eastern and western world can be found in these countries. Apparently, the sample of countries and regions to be used in my estimation is different from that used by, for example, Frankel et al. (1997) and Rauch (1999), both of the latter choose the cross-national data of the same 63 countries.5 The reason for my choice is as the following. First, I intend to present the full pictures for the US and Chinese foreign trade with the rest of the world (not just with their major trade partners only). Second, my panel data that include only the country (region) pairs connected with the U.S. or China will make the preparation of the cross-national data on cultural variables much easier. 6 The last section concludes with a brief discussion of policy implications to international traders.