• 沒有找到結果。

貿易流量之研究-以台灣製造業為例

N/A
N/A
Protected

Academic year: 2021

Share "貿易流量之研究-以台灣製造業為例"

Copied!
59
0
0

加載中.... (立即查看全文)

全文

(1)國立高雄大學應用經濟學系 碩士論文. 貿易流量之研究-以台灣製造業為例 On the Pattern of Bilateral Trade: The Selected Manufacturing Industries in Taiwan. 研究生:王珮華 撰 指導教授:鄭義暉 博士. 中華民國一百年七月. i.

(2) 致謝詞 在就讀研究所的兩年期間,多虧了許多人的幫忙,才能讓我現在順利修完學 分並且完成論文,當然,能夠順利畢業,在此首先要感謝 鄭義暉老師的耐心指 導和細心的教學,指導我撰寫論文的寫作方式和研究的方法,即便遇到困難的時 候,老師都還是花費許多的時間幫我解惑。即使在周末的時間,老師也都還是願 意犧牲假期撥空替我修改論文。 接下來要感謝姿妤姐在學校行政上程序的指導,還有 許聖章老師總是能夠 在我感到疲憊的時候和我促膝長談,好讓我有更多的動力繼續專注在課業上。另 外還要感謝 佘志民老師常常都讓哈利去老師的研究室吹冷氣休息。,以及 宋皇 叡老師熱情相挺常常借我腳踏車讓我可以輕輕鬆鬆去行政大樓跑公文,還有我親 愛的同學們,鳳儀、正宜以及衣瑄好友們常常都在日常生活中幫我許多忙,替我 分擔解憂,分享心情。 更感謝我的家人們給我很多精神上的支持,胚乳還有小莎莎常常陪我去戶外 踏青,還有和哈利、球球、多多一起去"跑跑",羅胖丁和喬治提供許多關於撰寫 論文的中肯意見。以及爸爸和媽媽給我很多關心和照顧,這些都是我進步的原動 力! 最後要特別感謝柯秀欣老師,若沒有柯老師的指導,我想今天的我沒有辦法 這麼順利的畢業,感謝柯老師在我人生的低潮中伸手拉了我一把,讓我學習到更 多待人處事的道理。最後的最後,我要感謝所有這兩年來幫忙我的所有朋友,在 研究所的這兩年,感謝你們的支持和照顧。謝天謝地,我終於畢業了!!嘻嘻!!. ii.

(3) 貿易流量之研究-以台灣製造業為例 指導教授:鄭義暉 博士 國立高雄大學應用經濟學系. 學生:王珮華 國立高雄大學應用經濟學系碩士班. 摘要 區域經濟整合為國際間貿易的新趨勢,各國透過簽訂自由貿易的 相關協定,冀藉以消除大部分貿易障礙及非貿易障礙,而增進會員國 間的貿易量,降低彼此間之交易成本,進而提升一國之社會福利。本 研究主要利用 1985-2009 年台灣與其主要貿易國之貿易資料,以四種 常見類型的貿易引力模型來探討雙邊貿易流量之差異。在區域經濟整 合體虛擬變數方面,本文選取現今主要的區域貿易協議,包括北美自 由貿易區、歐盟、歐元區、東南亞自由貿易區及南方共同市場,並加 入台灣前十大製造業產值的變數加以分析。本研究中使用模型包括: 跨部門綜合資料(FEIT)模型、三邊固定效果(FE3)、雙邊固定效 果(FE2)模型和四邊固定效果(FE4)模型,實證結果顯示若台灣 的貿易對手國若為 EU 的會員國則對台灣的出口量有正向且顯著的關 係。然而,若貿易對手國屬於 APEC 的會員國,則在部分模型中有負 向的效果存在。. 關鍵字:雙邊貿易、引力模型、區域經濟整合、製造業。 iii.

(4) On the Pattern of Bilateral Trade: Manufacturing Industries in Taiwan Advisor: Dr. I-Hui Cheng Department of Applied Economics National University of Kaohsiung Student: Pei-Hua Wang Department of Applied Economics National University of Kaohsiung. Abstract The main purpose of this study is to investigate the impact of regional economic agreement on bilateral trade flows, using various types of the gravity model of trade. For the comparison, the study focuses on the most commonly included regional trade agreement (REAs) in the literature, i.e. EU, NAFTA, AFTA, APEC and MERCOSUR. By comparing the results from estimating the modified conventional (FEIT), three-way fixed effects (FE3), two-way fixed effects (FE2) and four-way fixed effects (FEG) models, and judging by the use of manufacturing industries output and the trade volume of the industries. From the results of FE gravity model, the intra-trade impact of EU is positive and statistically significant. However, the intra-trade effect of APEC does negative effect under part of the specifications. Keywords:Bilateral trade, Gravity model, Regional economic agreement, manufacturing industries. iv.

(5) Contents. I.. Introduction ............................................................................................................ 1. 1. Motivation ........................................................................................................ 1. 2. Primary objective .............................................................................................. 4. II.. The literature ........................................................................................................ 6. III.. Contemporary world trade .............................................................................. 11. IV.. Data and estimation models ............................................................................. 23. 1. The data .......................................................................................................... 23. 2. The conventional models ................................................................................ 25. V.. The Estimation results ............................................................................. 36 1. The results of the FEIT and FECI gravity models ...................................... 37 2. The results of the two-way FE, three-way FE, and four-way FE models ..... 40. VI.. Conclusions ............................................................................................. 47. Data Appendix ............................................................................................................ 50 References ................................................................................................................... 51. v.

(6) List of tables Table 1.:. The total value of exports in different regions ................................................... 16. Table 2.:. The total value of imports in different regions ................................................... 17. Table 3.:. Taiwan’s trade structure by exporter partner ..................................................... 18. Table 4.:. Taiwan’s trade structure by importer partner ..................................................... 19. Table 5.:. Manufacturing export structure by partner of Taiwan ....................................... 20. Table 6.:. Manufacturing import structure by partner of Taiwan ....................................... 21. Table 7.:. Manufacturing output structure of Taiwan ........................................................ 22. Table 8.:. Variables and data sources .................................................................................. 33. Table 9.:. List of partner countries of Taiwan .................................................................... 34. Table 10.: Trade blocs and selected member countries ........................................................ 35. Table 11.: Regression results of various models ................................................................... 42. Table 12.: Regression results of FE models .......................................................................... 43. Table 13.: Regression results of FE models .......................................................................... 44. Table 14.: Regression results of various models ................................................................... 45. Table 15.: Regression results of various models ................................................................... 46. vi.

(7) I. Introduction 1.1. Motivation During the past twenty years, the growth of preferential trading agreements (PTAs) has attracted much attention. By forming PTAs, many countries eliminate tariffs, or lower trade barriers against their trading partners. There are various forms of PTAs, e.g., free trade area (FTA), customs union (CU), and common market and monetary union, which are classified according to the degree of economic integration among the members. The main purpose of preferential trading agreements is to lower the trade barriers between parties and to make members’ bilateral trade easier and more convenient. The cold war ended in the 1980s. Since then, economic activities have become more prosperous in homogeneous cultural environments than in heterogeneously ones. In addition, relationships among countries have tightened, the global competition has increased. Each country now looks for its own partner of trade, for example, by signing Preferential Trade Agreements (PTA) and joining Regional Economic Integration (REI). More examples can be discovered in research papers. For example, ASEAN Free Trade Area (AFTA) is one of the most famous trading groups in Southeast Asia. Another well-known treaty is, the North America Free Trade Agreement (NAFTA) 1.

(8) which includes America, Canada, and Mexico. In Africa, there is the Economic and Monetary Community of Central Africa (CEMAC). These trade agreements all represent some degree of economic integration in the areas in which they apply. This indicates, that most of the countries that have entered into these agreements do intend to integrate their markets with each other. Regional trade agreements have burgeoned in the last decade or so. While this can be good news, it can also be bad news. Such agreements are good when they bring regions closer together, create new profitable trading opportunities and set the scene for more inclusive market-opening. They are bad news when they discriminate unduly against third parties and frustrate the attainment of multilateral objectives built on non-discrimination. For the past several decades, the gravity model has been widely used to interpret the trade volume or the variations within countries and regions. The model is first initiated by Tinbergen and Pöyhönen in 1962 used to predict trade shares (Jeffrey Frankel, David Romer and Teresa Cyrus 1996). It also has become one of the most essential methods of explaining how national incomes and distances between two countries can influence their bilateral trade. In the mean time, social welfare is another vital issue that requires discussion. Recent decades have made it obvious that the volume of international trade plays a 2.

(9) significant role in determining a country’s GDP deeply affects a country’s national social welfare. Krishna (2005) discusses the design of welfare-improving PTAs that ensure gains for member countries without generating negative impacts on other countries. In another study, Frankel et al. (1994) points out that GNP per capita has a positive effect on trade: as countries develop, they tend to acquire more skills and hence specialize in more trades. These results all indicate that international trade and PTAs somehow help increase a country’s social welfare and development. By using the gravity model and some additional variables, this paper intends to assess the PTA’s effect on regional economies and the benefits obtained by both countries. The gravity model is chosen as the main measurement because it is widely used to explain international flows and labor migrations, and it has a strong theoretical foundation. In addition, we introduce several new variables in this study, including values of output in manufacturing industries. .. 3.

(10) 2. Primary objective The purpose of this study is to use the gravity model to find out whether the signing of REAs can bring certain effects to a country’s trade flows and effects to a country’s bilateral trade to other regions or areas. The remainder of this study is organized as follows. Chapter two will briefly reviews the literature on the gravity model, bilateral trade, PTAs, and regional economies. Chapter three outlines the current condition of world trade, focusing on economic conditions beween Taiwan and its trading partners. Chapter four discusses the data, and Section five provides an empirical analysis. Finally, Chapter six offers some concluding remarks.. 4.

(11) II. The Literature Among various types of empirical trade models, the gravity model is one of the most widely used in the study of bilateral trade. The foundation of the gravity model was basically established on the Sir Isaac Newton’s Law of Universal Gravitation. According to the Law, there exists gravity between any two objects, and the degree of the gravitation depends on the distance between these two objects, their quality and quantity. The gravity model was firstly introduced in social sciences and popularly used in 1940s. Its empirical robustness makes it suitable for the study on the geographical patterns of migration and transportation, etc. Tinbergen (1962) and Pöyhönen (1963) provide the initial specifications of the gravity-type model to estimate the determinants of bilateral trade flows. In the basic form of the gravity model of trade, the volume of trade flows between two countries is assumed to be increasing in their sizes, as measured by their national incomes, and decreasing in the cost of transportation, as measured by the distance between their economic centers. Linnemann (1966) adds population as a measurement of country size, suggesting that populations of a trading-pair countries have a significantly negative effect on trade flows. Anderson (1979) considers the properties of the expenditure system in which two cases of product differentiation are discussed, namely the Cobb-Douglas (C-D) and constant-elasticity-of-substitution (CES) 5.

(12) preferences, to derive the theoretical gravity model. Bergstrand (1985) includes the variables of prices, and adopts the CES preferences to derive the gravity equation. Bergstrand (1989) further developes a general model of world in which with two differentiated-product industries are considered. In the study, it is demonstrated that the gravity equation, including exporter and importer populations as well as incomes, can well fit in with the Heckscher-Ohlin model of inter-industry trade and the Helpman-Krugman-Markusen model of intra-industry trade. Deardorff (1988) builds bilateral trade model with the Hecksher-Ohlin structure, considering the frictionless trade and impeded trade cases. In the study, it shows that in the presence of transportation costs and tariffs, the pattern of trade flows on a C.I.F. basis will be the same with the frictionless trade case. With the C-D preference, the gravity equation of trade flows on an F.O.B. basis is illustrated, weighted with the iceberg-form transportation costs. With the CES preference, the measurement of the value of trade flows on an F.O.B. basis is more complicated. The pattern of trade flows can be derived as a function of the incomes of importing and exporting countries, exporter’s share of world income, the relative and average distances from suppliers.. 6.

(13) Anderson & Wincoop (2003) develop a method to consistently and efficiently estimate a theoretical gravity equation, and apply the method to solve the border puzzle. The study points out that natural border reduces bilateral trade flows across countries by plausible though substantial magnitudes. Krugman (1991) derives a model in which every regional bloc pursues an optimal tariff. It shows that three regional blocs may minimize world welfare. In a generalized version of Krugman’s model that includes transport costs, Frankel et al (1998) show that even taking the geographic pattern of trading blocs into account and justifying a certain degree of regional preferences as “natural”, the degree of regionalization is likely to be excessive and expected to be welfare-reducing. The study of Guo (2004) illustrates that using common language makes traders to communicate with each other easier and more convenient, and thus lowers their transaction costs. The study also shows that both religion and culture have a positive effect on their bilateral trade. Frankel & Wei (1993) use the dummy variable to differentiate between the regions which have a common language or not, and argue that sharing the same language has a significant effect on bilateral trade. Regional economic integration allows the process of eliminating the trade barriers among member countries, reducing their limit to the shift of production factors, and expanding markets of product and service. That is, regional economic 7.

(14) integration could expand market scale, encourage international competition, and thus can possibly benefit all members in the process.. Aitken (1973) utilizes a cross-sectional trade flow model of the gravity-type to study the impacts of the European Economic Community (EEC) and the European Free Trade Association (EFTA) on member country’s trade. The results show that both the EEC and EFTA have experienced a cumulative growth in gross trade creation (GTC) over their respective integration period. Frankel et al. (1995) use the gravity model to examine bilateral trade patterns, arguing that world trade is more regionalized. Intra-regional trade is greater than could be explained by so-called natural determinants, e.g. proximity of a pair of countries, their sizes and GNP per capitas, and whether they share a common border or a common language.. Oguledo et al. (1994) estimate a reformulated gravity model on bilateral trade flows among the more developed countries. It shows that the tariff rate does not fully reflect all the factors which influence the flow of trade between preference-giving and preference-receiving countries. Soloaga & Winters (2001) use a gravity model to quantify the effects of PTAs on trade, using a data set of 58 non-fuel import countries over Year 1980 to Year 1996. The study also provides evidence of export diversion in the EU and EFTA.. 8.

(15) Bayoumi & Eichengreen (1997) find that the formation of the EEC and EFTA had significant impacts on Europe’s trade during 1958-1980. However, it cannot be attributed to participating countries’ observable economic characteristic or unobservable factors, such as historical trade relations or preferential trade structures. Cheng & Tsai (2005) conclude that the exporter’s and importer’s incomes have positive effects on bilateral trade flows, but the income elasticity of bilateral trade is less than that estimated by using the conventional gravity specification. However, the coefficients of the exporter’s and importer’s populations are positive and statistically significant, which is contrary to the conventional wisdom.. Bergstrand & Baier (2005) argue that treating trade policy as an exogenous variable to estimate whether FTAs actually do increase members’ international trade. After forty years of gravity equation estimates of the effect of FTAs on trade flows, FTAs will on average increase bilateral trade among two member countries. Furthermore, Baier & Bergstrand (2008) provide the first cross-section estimates of long-run treatments effects of free trade agreements on members’ bilateral international trade flows by using matching econometrics. Their nonparametric cross-section estimates of ex post long-run treatment effects are more stable across years.. 9.

(16) According to Egger & Larch (2007), there is a robust implication of a positive interdependence in PTA membership in the world economy. The formation of PTAs generates a particularly strong incentive for non-distant outsiders to join, and there is a somewhat lower but still positive incentive for outsiders to form their own PTA. Egger et al. (2006) formulate an empirical model to estimate the impact of new RTA membership on trade structure. The likelihood of new RTA membership is influenced by economic fundamentals such as country size, factor endowments and trade and investment costs. Overall, RTA membership might reduce inter-industry trade not only in relative but also in absolute terms. It is shown that trade volume effect is due to the associated growth in trade within industries.. 10.

(17) III. Contemporary World Trade In last two decades, the values of imports and exports have increased significantly across different geographical regions. As shown in Table 1, exports in Eastern Asia have experienced significant growth during Year 1980 to Year 2009, increasing from 76,165 million dollars to 2,090,019 million dollars, while its export share of world trade significantly rises from 3.74 percent to 15.7 percent. During the same period, exports in developing South-Eastern Asian countries increased from 73,957 million dollars to 811,750 million dollars, with the share of world trade rising from 3.63 percent to 6.54 percent.. The strong expansion of developing Asia’s trade evidently reveals the continued expansion of South-South trade. According to Asian Development Bank annual report, this is because developing Asia trades proportionally more with developing member countries than other developing non-member regions. In contrast, the values of exports of developed European countries increased from 892,271 million dollars to 4,851,322 million dollars during Year 1980 to Year 2009, however, its share of exports relative to the world trade here decreased from 43.85 percent to 39.06 percent.. The total value of imports in different regions in the last two decades is summarized in Table 2. The values of imports in developed America have experienced evident growth during Year 1980 to Year 2009, from 320,210 million dollars to 11.

(18) 1,937,024 million dollars, while the share of world trade remains roughly the same from 15.42 percent to 15.39 percent. During the same period, the total imports values in Europe rises up from 998,105 million dollars to 4,833,112 million dollars, whereas the share of trade significantly decreased from 48.06 percent to 38.39 percent. After 1992, total volumes of exports in Developed Europe increased from 1,647,000 million dollars to 2,314,897 million dollars from Year 1990 to Year 1995, while total volumes of imports in the region increased from 1,750,494 million dollars to 2,241,443 million of dollars.. Taiwan is one of the most prosperous economies in East Asian. In the past ten years, the international trade flows in Taiwan have grown significantly especially after Year 1959, when Taiwan’s economy started to grow rapidly due to the government policy. As illustrated in Table 3 and Table 4, the total values of exports of Taiwan to China increased from 377 million of dollars to 66,564 million of dollars, whereas the share of exports from Taiwan to China rose up from 0.33 percent to 26.04 percent. During the same period, the total values of imports of Taiwan from China increased from 3,088 million of dollars to 31,451 million of dollars, while the share of imports significantly increases from 2.97 percent to 13.08 percent.. Furthermore, the total exports values of Taiwan to Korea have growth significantly from 2,564 million of dollars and 2.26 percent to 7,296 million of dollars 12.

(19) and 3.58 percent, and the total values of Taiwan’s imports from Korea increased from 4,327 million of dollars and 4.16 percent to 10,530 million of dollars and 6.04 percent. From Year 1995 to Year 2002, the total values of exports of Taiwan to Hong Kong rose up from 26,012 million of dollars and 22.95 percent to 30,847 million of dollars and 22.80 percent. From Year 2003 to Tear 2008, the values of exports increased from 28,273 million of dollars to 32,603 million of dollars, while unit of percentage fell from 18.77 percent to 12.75 percent.. As shown in Table 5, the values of Taiwan’s manufacturing exports to China increased from 5,644 million dollars to 82,732 million dollars from Year 1995 to Year 2009, whereas the share of manufacturing exports significantly increased from 19.18 percent to 70.48 percent. During the same period, the total values of manufacturing exports from Taiwan to Japan rose up from 111 million dollars to 1,284 million dollars, while unit of percentage of manufacturing exports increased from 0.38 percent to 1.09 percent, which have grown about triple times in the same period.. Furthermore, the total values of manufacturing exports increased from 6 million of dollars to 498 million of dollars from Year 1995 to Year 2009 in Korea, the values of manufacturing exports have significantly increased about 83 times, while the percentage of manufacturing exports rose up from 0.02 percent to 0.42 percent. It reveals that trade between Taiwan and Korea has become more and more prosperous. 13.

(20) As seen in Table 6, the values of Taiwan’s manufacturing imports from China increased from 2,586 million of dollars to 23,162 million of dollars from Year 1995 to Year 2009 in Taiwan, while the unit of percentage fast up from 2.85 percent to 17.16 percent. In India, the values of manufacturing imports has risen approximately four times, from 344 million of dollars in Year 1995 to 1,488 million of dollars in Year 2009, whereas the share of manufacturing exports has risen from 0.38 percent to 1.10 percent during the same period. The total values of manufacturing imports of Taiwan form Japan fast up from 30,070 million of dollars in Year 1995 to 35,713 million of dollars in Year 2009, while the share of manufacturing imports has decreased from 33.10 percent to 26.46 percent, it seems that the values of trade between Taiwan and Japan has increased whereas the share of Taiwan’s import from Japan actually reduced about 6 percent.. As illustrated in Table 7, the values of output of manufacture of coke and refined petroleum products fast up from 7,124 million of dollars in Year 1995 to 28,842 million of dollars in Year 2009, while the share has increased from 0.13 percent to 0.26 percent. during Year 1995 to Year 2009, the share of output of manufacture of computer, electronic and optical products has risen from 0.56 percent in Year 1995 to 0.75 percent in Year 2009, while the values of output has increased to 82,957 million of dollars from 30,796 million of dollars. Also in Table 7, the share of output of 14.

(21) manufacture of chemicals and chemical products has increased, while the share of output of manufacture of rubber and plastic products and manufacture of motor vehicles, trailers and semi-trailers have declined.. 15.

(22) Table 1. The total value of exports in different regions YEAR. 1980. 1990. 1995. 2000. 2005. Millions of dollars 2006. 2007. 2008. (Percentage) 2009. Developing economies 121,876 Africa (5.99%) 111,352 America (5.47%). 110,670 (3.18%) 143,911 (4.13%). 114,221 (2.21%) 229,435 (4.43%). 149,375 (2.36%) 366,509 (5.65%). 303,036 (2.89%) 570,768 (5.45%). 373,125 (3.07%) 576,659 (5.66%). 436,454 (3.07%) 773,256 (5.52%). 567,048 (3.45%) 900,860 (5.52%). 383,631 (3.09%) 687,216 (5.53%). 76,165 (3.74%) 73,957 (3.63%). 280,565 (8.05%) 145,284 (4.17%). 562,608 (10.86%) 323,454 (6.25%). 774,898 (12.02%) 431,944 (6.70%). 1,538,367 (14.69) 652,733 (6.23%). 1,846,744 (15.23%) 770,656 (6.35%). 2,186,288 (15.67%) 864,964 (6.20%). 2,475,315 (15.45%) 1,003,790 (6.26%). 2,090,019 (16.83%) 811,750 (6.54%). 521,758. 777,365. 1,058,871. 1,267,055. 1,428,252. 1,580,115. 1,753,793. 1,373,179. (14.98%) 299,157 (7.01%) 1,647,000 (47.28%). (15.01%) 462,162 (6.98%) 2,314,897 (44.70%). (16.42%) 510,653 (6.26%) 2,591,769 (40.19%). (12.1%) 637,675 (6.09%) 4,307,508 (41.12%). (11.78%) 696,720 (5.09%) 4,860,295 (40.07%). (11.32%) 754,603 (4.77%) 5,647,561 (40.46%). (10.94%) 847,255 (5.08%) 6,247,494 (38.98%). (11.06%) 628,780 (5.06%) 4,851,322 (39.06%). Eastern Asia South-Eastern Asia. Developed economies 293,549 America (14.43%) 135,979 Asia (7.00%) 892,271 Europe (43.85%). Sources: UNCTAD Handbook of Statistics 2010, UN Yearbook of International Trade Statistics, IMF Direction of Trade Statistics. Note: Exports are calculated in F.O.B unit and measured in millions of dollars. The shares of exports relative to the world trade in different regions are in parentheses. Hsieh (2009.). 16.

(23) Table 2. The total value of imports in different regions YEAR. 1980. 1990. 1995. 2000. 2005. Millions of dollars 2006. 2007. (Percentage) 2008. 2009. Developing economies Africa America East Asia South-Eastern Asia. 96,856. 101,718. 126,614. 132,878. 262,927. 312,189. 378,503. 482,155. 414,534. (4.66%). (2.83%). (2.42%). (1.99%). (2.44%). (2.52%). (2.66%). (2.95%). (3.30%). 123,594. 127,275. 250,711. 392,134. 530,415. 633,841. 753,712. 902,341. 683,954. (5.95%). (3.54%). (4.79%). (5.89%). (4.92%). (5.12%). (5.30%). (5.52%). (5.43%). 85,536. 265,896. 567,473. 742,868. 1,411,111. 1,647,693. 1,910,483. 2,210,617. 1,857,709. (4.12%). (7.40%). (10.83%). (11.15%). (13.09%). (13.30%). (13.43%). (13.53%). (14.76%). 65,641. 162,292. 355,323. 377,441. 594,331. 687,976. 776,356. 948,366. 722,896. (3.16%). (4.51%). (6.78%). (5.67%). (5.52%). (5.55%). (5.46%). (5.80%). (5.70%). 320,210. 641,358. 939,911. 1,505,270. 2,065,547. 2,278,836. 2,408,102. 2,584,666. 1,937,024. (15.42%). (17.84%). (17.94%). (22.59%). (19.18%). (18.40%). (16.92%). (15.82%). (15.39%). 151,080. 252,162. 365,461. 417,197. 562,064. 629,908. 678,868. 830,248. 599,831. (7.27%). (7.01%). (6.98%). (6.26%). (5.22%). (5.09%). (4.77%). (5.08%). (4.76%). 998,105. 1,705,494. 2,241,443. 2,630,452. 4,327,522. 4,947,006. 5,750,000. 6,396,748. 4,833,112. (48.06%). (47.44%). (42.78%). (39.48%). (40.16%). (39.93%). (40.41%). (39.15%). (38.39%). Developed economies America Asia Europe. Sources: UNCTAD Handbook of Statistics 2010, UN Yearbook of International Trade Statistics, IMF Direction of Trade Statistics. Note: Imports are calculated in C.I.F. unit and measured in millions of dollars. The shares of imports relative to the world trade in different regions are in parentheses. Hsieh (2009.). 17.

(24) Table 3. Taiwan’s trade structure by exporter partner DESTINATION TOTAL China. YEAR. Hong Kong India Indonesia Japan Korea Malaysia Philippines Singapore Thailand Viet Nam. 1995 55,600 377 (0.33) 26,012 (22.95) 519 (0.46) .. .. 13,122 (11.58) 2,564 (2.26) 2,890 (2.55) 1,649 (1.45) 4,396 (3.88) 3,061 (2.70) 1,010 (0.89). 2000 73,101 4,217 (2.78) 31,335 (20.62) 717 (0.47) .. .. 16,599 (10.92) 3,907 (2.57) 3,611 (2.38) 3,035 (2.00) 5,455 (3.59) 2,562 (1.69) 1,663 (1.09). 2001 61,484 4,745 (3.76) 26,964 (21.35) 630 (0.49) .. .. 12,759 (10.10) 3,275 (2.60) 3,061 (2.42) 2,148 (1.70) 4,051 (3.20) 2,125 (1.68) 1,726 (1.37). 2002 71,353 9,950 (7.35) 30,847 (22.80) 647 (0.48) .. .. 11,983 (8.86) 3,866 (2.86) 3,132 (2.31) 1,971 (1.46) 4,377 (3.23) 2,293 (1.70) 2,287 (1.70). Sources: UNCTAD Handbook of Statistics 2010.. 18. 2003 83,985 21,405 (14.21) 28,273 (18.77) 768 (0.51) 1,511 (1.00) 11,881 (7.89) 4,569 (3.03) 3,041 (2.02) 2,297 (1.53) 4,976 (3.30) 2,559 (1.70) 2,660 (1.77). 2004 106,088 33,997 (18.64) 29,727 (16.30) 1,067 (0.59) 1,863 (1.02) 13,161 (7.22) 5,345 (2.93) 4,067 (2.23) 3,891 (2.13) 6,333 (3.47) 3,212 (1.76) 3,425 (1.88). 2005 119,365 40,879 (20.6) 30,721 (15.49) 1,568 (0.79) 2,336 (1.18) 14,481 (7.30) 5,575 (2.80) 4,154 (2.09) 4,220 (2.13) 7,656 (3.86) 3,718 (1.87) 4,057 (2.04). Millions of dollars (Percentage) of dollars 2006 2007Millions 2008 2009 114,762 162,040 166,715 138,018 51,808 62,360 66,564 54,163 (23.13) (25.28) (26.04) (26.59) 37,380 37,933 32,603 29,429 (16.69) (15.38) (12.75) (14.45) 1,471 2,340 2,992 2,534 (0.66) (0.95) (1.17) (1.24) 2,499 2,905 3,561 3,217 (1.12) (1.18) (1.39) (1.58) 16,300 15,908 17,564 14,498 (7.28) (6.45) (6.87) (7.12) 7,154 7,779 8,679 7,296 (3.19) (3.15) (3.40) (3.58) 4,941 5,382 5,496 4,050 (2.20) (2.18) (2.15) (1.99) 4,484 4,914 4,774 4,427 (2.00) (1.99) (1.87) (2.17) 9,279 10,480 11,662 8,592 (4.14) (4.25) (4.56) (4.22) 4,577 5,193 4,898 3,822 (2.04) (2.11) (1.92) (1.88) 4,869 6,846 7,922 5,990 (2.17) (2.78) (3.10) (2.94).

(25) Table 4.. SOURCE TOTAL China. YEAR. Hong Kong India Indonesia Japan Korea Malaysia Philippines Singapore Thailand Viet Nam. Taiwan’s trade structure by importer partner 1995 2000 2001 2002 2003 2004 48,216 73,630 54,220 59,185 71,766 94,826 3,088 6,223 5,901 7,947 10,917 16,625 (2.97) (4.42) (5.47) (7.02) (8.53) (9.85) 1,842 2,185 1,847 1,738 1,701 2,072 (1.77) (1.55) (1.71) (1.53) (1.33) (1.23) 413 513 493 550 622 858 (0.40) (0.36) (0.46) (0.49) (0.49) (0.51) .. .. .. .. 2,917 4,106 .. .. .. .. (0.49) (0.51) 30,250 38,556 25,847 27,276 32,513 43,496 (29.08) (27.40) (23.94) (24.09) (25.40) (25.77) 4,327 8,987 6,704 7,711 8,673 11,610 (4.16) (6.39) (6.21) (6.81) (6.78) (6.88) 2,955 5,325 4,213 4,151 4,736 5,393 (2.84) (3.78) (3.90) (3.67) (3.70) (3.20) 624 3,593 3250 3,651 3,074 3,048 (0.60) (2.55) (3.01) (3.22) (2.40) (2.53) 2,962 5,013 3,367 3,543 3,810 4,267 (2.85) (3.56) (3.12) (3.13) (2.98) (2.53) 1,485 2,767 2,180 2,170 2,353 2,753 (1.43) (1.97) (2.02) (1.92) (1.84) (1.63) 270 468 418 448 450 598 (0.26) (0.33) (0.39) (0.40) (0.35) (0.35). Sources: UNCTAD Handbook of Statistics 2010. 19. Millions of dollars (Percentage) 2005 2006 2007 Millions 2008of dollars 2009 102,829 112,439 117,380 120,702 94,036 19,928 24,728 28,058 31,451 24,491 (10.91) (12.20) (12.80) (13.08) (14.05) 1,887 1,880 1,828 1,490 1,123 (1.03) (0.93) (0.83) (0.62) (0.64) 857 1,245 2,543 2,329 1,625 (0.47) (0.61) (1.16) (0.97) (0.93) 4,538 5,204 5,789 7,308 5,215 (2.49) (2.57) (2.64) (3.04) (3.00) 45,940 46,284 46,017 46,622 36,313 (25.16) (22.83) (20.99) (19.39) (20.83) 13,203 14,999 15,184 13,189 10,530 (7.23) (7.40) (6.93) (5.49) (6.04) 5,194 6,052 6,215 6,768 4,696 (2.84) (2.99) (2.83) (2.81) (2.70) 2,786 2,776 2,282 2,242 1,618 (1.53) (1.37) (1.04) (0.93) (0.93) 4,940 5,104 4,800 4,825 4,811 (2.71) (2.52) (2.19) (2.01) (1.75) 2,867 3,317 3,620 3,262 2,690 (1.57) (1.64) (1.65) (1.36) (1.54) 689 850 1,044 1,216 924 (0.38) (0.42) (0.48) (0.51) (0.53).

(26) Table 5.. DESTINATION TOTAL China Hong Kong India Indonesia Japan Korea Malaysia Philippines Singapore Thailand Viet Nam. Manufacturing export structure by partner of Taiwan Millions of current dollars (Percentage) YEAR 1995 2000 2001 2002 2003 2004 2005 2006 2007 Millions 2008of dollars 2009 9,586 24,768 28,355 35,489 44,236 52,440 58,867 60,237 70,200 92,913 100,330 5,644 17,103 19,887 26,610 34,309 41,249 47,256 47,256 54,899 75,560 82,730 (19.18) (31.99) (33.74) (41.69) (49.36) (53.81) (60.94) (57.15) (60.94) (67.78) (70.48) 696 1,114 1,209 1,376 2,017 2,157 2,265 2,537 2,726 3,064 3,305 (2.37) (2.08) (2.05) (2.16) (2.90) (2.81) (2.74) (2.82) (2.69) (2.75) (2.82) 7 8 11 14 15 16 18 22 30 46 49 (0.02) (0.01) (0.02) (0.02) (0.02) (0.02) (0.02) (0.02) (0.03) (0.04) (0.04) 346 545 551 561 573 576 585 594 594 597 599 (1.18) (1.02) (0.93) (0.88) (0.82) (0.75) (0.70) (0.66) (0.59) (0.54) (0.51) 111 614 783 807 907 1,056 1,099 1,110 1,129 1,181 1,284 (0.38) (1.15) (1.33) (1.26) (1.30) (1.38) (1.33) (1.23) (1.12) (1.06) (1.09) 6 188 201 206 216 223 226 242 253 489 498 (0.02) (0.35) (0.34) (0.32) (0.31) (0.29) (0.27) (0.27) (0.25) (0.44) (0.42) 1,183 1,414 1,460 1,492 1,542 1,577 1,606 1,637 1,702 1,730 1,831 (4.02) (2.64) (2.48) (2.34) (2.22) (2.06) (1.94) (1.82) (1.68) (1.55) (1.55) 283 566 612 695 697 699 714 728 741 774 766 (0.96) (1.06) (1.04) (1.09) (1.00) (0.91) (0.86) (0.81) (0.73) (0.67) (0.65) 284 1,382 1,760 1,786 1,812 2,635 2,732 3,539 4,733 5,430 5,467 (0.97) (2.58) (2.99) (2.80) (2.61) (3.44) (3.30) (3.93) (4.68) (4.87) (4.66) 614 1,037 1,053 1,059 1,108 1,117 1,137 1,219 1,931 1,940 1,955 (2.09) (1.94) (1.79) (1.66) (1.59) (1.46) (1.38) (1.35) (1.91) (1.74) (1.67) 412 797 828 883 1,040 1,135 1,229 1,353 1,462 2,102 2,344 (1.40) (1.49) (1.40) (1.38) (1.50) (1.48) (1.49) (1.50) (1.44) (1.89) (1.20). Sources: UNCTAD Handbook of Statistics 2010.. 20.

(27) Table 6.. DESTINATION TOTAL China Hong Kong India Indonesia Japan Korea Malaysia Philippines Singapore Thailand Viet Nam. Manufacturing import structure by partner of Taiwan YEAR 1995 2000 2001 2002 2003 2004 47,861 73,067 53,218 57,982 67,526 88,469 2,586 5,624 5,126 7,068 9,932 15,203 (2.85) (4.59) (5.61) (7.37) (9.28) (10.80) 1,809 2,082 1,753 1,682 1,662 2,021 (1.99) (1.70) (1.92) (1.75) (1.55) (1.44) 344 429 432 475 526 720 (0.38) (0.35) (0.47) (0.50) (0.49) (0.51) 1,153 1,599 1,197 1,141 1,308 1,604 (1.27) (1.30) (1.31) (1.19) (1.22) (1.14) 30,070 38,296 25,572 26,918 32,174 42,967 (33.10) (31.25) (28.00) (28.07) (30.07) (30.52) 4,261 8,914 6,652 7,641 8,621 11,511 (4.69) (7.27) (7.28) (7.97) (8.06) (8.18) 2,557 4,589 3,576 3,598 4,025 4,352 (2.82) (3.74) (3.92) (3.75) (3.76) (3.09) 604 3,562 3,205 3,582 2,955 2,812 (0.66) (2.91) (3.51) (3.74) (2.76) (2.00) 2,946 4,977 3,336 3,502 3,804 4,252 (3.24) (4.06) (3.65) (3.65) (3.56) (3.02) 1,300 2,602 2,017 1,985 2,131 2,525 (1.43) (2.12) (2.21) (2.07) (1.99) (1.79) 231 393 352 390 388 502 (0.25) (0.32) (0.38) (0.41) (0.36) (0.36). Sources: UNCTAD Handbook of Statistics 2010.. 21. Millions of current dollars (Percentage) 2005 2006 2007 Millions 2008of dollars 2009 95,350 103,831 108,205 108,427 86,105 18,168 22,875 26,051 28,678 23,162 (12.29) (14.14) (14.96) (16.11) (17.16) 1,800 1,718 1,550 1,148 966 (1.22) (1.06) (0.89) (0.65) (0.72) 766 1,152 2,430 1,983 1,488 (0.52) (0.71) (1.40) (1.11) (1.10) 1,567 1,808 2,040 2,034 1,407 (1.06) (1.11) (1.17) (1.14) (1.04) 45,228 45,458 45,249 45,827 35,713 (30.60) (28.09) (25.99) (25.75) (26.46) 13,073 14,839 14,951 13,005 10,375 (8.84) (9.17) (8.59) (7.31) (7.69) 4,079 4,554 4,813 4,830 3,361 (2.76) (2.81) (2.76) (2.71) (2.49) 2,560 2,570 2,134 2,075 1,548 (1.73) (1.59) (1.23) (1.17) (1.15) 4,873 5,065 4,767 4,781 4,777 (3.30) (3.13) (2.74) (2.69) (3.54) 2,655 3,074 3,358 3,021 2,518 (1.80) (1.90) (1.93) (1.70) (1.87) 581 718 862 1,045 790 (0.39) (0.44) (0.50) (0.59) (0.59).

(28) Table 7.. ISIC INDUSTRIAL CLASIFFICATION TOTAL 19 Manufacture of coke and refined petroleum products. YEAR. Manufacturing output structure of Taiwan 1995 2000 2001 2002 2003 2004. Millions of dollars (Percentage) Millions of dollars 2005 2006 2007 2008 2009. 123,979 171,985 136,175 150,678 175,755 227,446 253,752 280,746 315,037 325,523 243,859 7,124 12,724 12,432 12,484 15,465 20,721 27,380 30,721 33,874 40,022 28,842 (0.13) (0.17) (0.18) (0.17) (0.19) (0.23) (0.28) (0.29) (0.29) (0.34) (0.26). 20. Manufacture of chemicals and chemical products. 22. Manufacture of rubber and plastics products. 24. Manufacture of basic metals. 25. Manufacture of fabricated metal products, except machinery and equipment. 13,246 (0.24) 8,564 (0.15) 18,223 (0.33) 16,001 (0.29). 23,140 (0.32) 9029 (0.12) 21,700 (0.30) 16,985 (0.23). 26. Manufacture of computer, electronic and optical products. 30,796 54,706 38,713 45,394 52,608 66,731 75,974 91,036 100,581 98,550 82,957 (0.56) (0.75) (0.57) (0.61) (0.65) (0.74) (0.78) (0.86) (0.87) (0.85) (0.75). 27. Manufacture of electrical equipment. 28. Manufacture of machinery and equipment n.e.c.. 29. Manufacture of motor vehicles, trailers and semi-trailers. 10,005 (0.18) 10,893 (0.20) 9,127 (0.16). 11,027 (0.15) 13,172 (0.18) 9,502 (0.13). 21,675 (0.32) 7,449 (0.11) 16,270 (0.24) 13,631 (0.20). 8,628 (0.13) 10,361 (0.15) 7,016 (0.10). Sources: UNCTAD Handbook of Statistics 2010.. 22. 23,146 (0.31) 7,543 (0.10) 19,458 (0.26) 14,359 (0.19). 8,929 (0.12) 10,985 (0.15) 8,380 (0.11). 28,570 (0.35) 7,727 (0.09) 24.008 (0.29) 15,660 (0.19). 9,171 (0.11) 12,585 (0.15) 9,961 (0.12). 39,720 (0.44) 8,501 (0.09) 34,698 (0.39) 19,408 (0.22). 10,648 (0.45) 15,199 (0.17) 11,820 (0.13). 43,590 (0.45) 8,415 (0.09) 37,468 (0.39) 20,286 (0.21). 10,985 (0.11) 16,791 (0.17) 12,863 (0.13). 47,117 (0.45) 7,948 (0.08) 42,913 (0.41) 21,048 (0.20). 12,418 (0.12) 17,563 (0.17) 9,982 (0.09). 58,052 (0.50) 8,015 (0.07) 49,680 (0.43) 23,076 (0.20). 12,671 (0.11) 19,138 (0.16) 9,950 (0.09). 58,767 (0.50) 8,083 (0.07) 55,370 (0.47) 24,535 (0.21). 12,878 (0.11) 19,298 (0.17) 8,020 (0.07). 45,710 (0.41) 6,579 (0.06) 34,014 (0.31) 16,171 (0.15). 9,387 (0.08) 12,222 (0.11) 7,977 (0.07).

(29) Ⅳ. The Data and Empirical Models 4.1. The data. This study uses the F.O.B. values of Taiwan exports to its trading partners as the dependent variable. The independent variables include trade partners’ real GDP, population, distances from the trading partner to Taiwan, and the dummy variables to point out if a Taiwan’s trading partner is acting in any major regional bloc. The descriptions of these variables and data resources are provided in Table 8.. The dataset used in this research is a balanced panel with 5,550 observations with 37 Taiwan’s trading partners, using trading values of exports and imports of top 10 manufacturing industries from Year 1980 to 2009. Table 9 demonstrates the list of trading partners of Taiwan included in this study. To discuss the effect of regional trading agreements, this study employs five regional bloc dummy variables, namely the European Union (EU), North American Free Trade Agreement (NAFTA), Asia Pacific Economic Cooperation (APEC), Free Trade Area (AFTA) and MERCOSUR, descriptions of these dummies are shown in Table 10. In this study, a trading bloc dummy is set to be one if a partner country is belonging to the specific trading bloc in a given year. In Year 1950, the European Coal and Steel Community (ECSC) was established to unite economic power among. 23.

(30) member countries. European Economic Community (EEC) came into effect in Year 1957. The six founder of countries are Belgium, France, Germany, Italy, Luxembourg and the Netherlands, respectively. Afterwards, Denmark, Ireland and the United Kingdom joined the EEC in Year 1973, Greece in Year 1981, and Spain and Portugal in Year 1986. Czech Republic, Estonia, Latvia, Lithuania, Hungary, Poland, Slovenia and Slovakia then participated in Year 2004, and Bulgaria and Romania joined the EU in Year 2007. In this research, the dummy of EU is set to be one when partner countries of Taiwan are acting in EU in year t. The Canada-US Free Trade Agreement (CUFTA) was established in Year 1989. The agreement was extended to include Mexico, and form a new trading bloc named as North American Free Trade Agreement which entered into effect in 1994. To measure the effect, this study sets the NAFTA dummy to be one if a trading partner of Taiwan is the US, Canada, or Mexico, and otherwise zero in this study. The Asia Pacific Economic Cooperation (APEC) was formed in Year 1991 by the Australia, America, Canada, Japan, South Korea and New Zealand, etc. In this study, the AFTA dummy is set to be one if the trading partner of Taiwan is acting in the AFTA. The MERCOSUR was created by four South America countries, Argentina, Brazil, Paraguay and Uruguay in 1994. Then, Venezuela joined in 2006. To estimate the effect, this study lets the MERCOSUR dummy to be one if the trading 24.

(31) partner of Taiwan is acting in the MERCOSUR, and otherwise zero.. 4.2. The conventional models This study uses various forms of the gravity model to analyze the determination of trade flows. The gravity model of international trade assumes that the values of bilateral trade flows can be determined as an increasing function of the sizes of trading economies, a decreasing function of the geographic distance between them, and other economic and non-economic factors, which may enhance or deter bilateral trade. A simple form of the gravity model takes the following form: lnEXijkt=αt+αk+β1lnYit+β2lnYjt+β3lnDISTij+β4LANGij +β5ADJij+β6REAijt+εijkt ,. (CS). where EXijkt is the value of exports from country i to country j in year t, and αt is the year dummy, DISTij is the distance between economic centers of country i and country j, LANGij is a dummy variable assuming the value of one if two countries have a common official language, and otherwise zero, ADJij is a dummy variable assuming that the value of one if two countries share a common border and zero otherwise, and REAijt is a dummy variable assuming the value of one if the country join a regional economic agreement in year t and zero otherwise, εijkt is the residual term. The CS model is usually estimated by the ordinary least squares (OLS) method. 25.

(32) This study uses pooled cross-section (PCS) model as a benchmark to measure the determinants of bilateral trade flows. The PCS estimation of this study is referred to Brada & Mendez (1983) in which a cross-section approach is used to study trade flows. In the following, I use the following four types of gravity model to explain the determinants of the bilateral trade: (1) controlling for the year (αt) and industry-specific effects (αk) (Model FEIT), and (2) controlling for the year (αt) and country-industry pair effects (αjk) (Model FECI), and (3) controlling for the year (αt), country (αj), and industry specific effects (αk) (Model FE3), and (4) controlling for the year (αt), country (αj), industry (αk), and country-industry pair effects (αjk) (Model FEFV).. Controlling for the year and industry specific effects (Model FEIT) Undoubtedly, pooling cross-section and time-series data increases the number of observations for the estimation. The pooled cross-section model (PCS) deals with the problem of additional degree of freedom without significantly increasing the number of variables, allowing for the intercepts to differ over time. To measure the effects of trade blocs on bilateral trading pattern, the gravity model could be specified as: lnEXijkt=αt +αk +β1lnYjt+β2lnNjt+β3lnFDI_OUTjt+β4lnFDI_INjt +β5lnKjt+β6lnMANUjt+β7lnMANU_LABOjt+β8lnDISTij 26.

(33) +β9lnINDUS_DISTijkt+β10EUjt+β11NAFTAjt+β12AFTAjt +β13MERCOSURjt+β14APECjt+εijkt ,. (FEIT). where EXijkt is the value of industry exports from country i to country j in year t and αt is the year-specific effect common to Taiwan and other trading countries, αk is the industry-specific effect, and FDI_OUTjt is the value of outward of foreign direct investment of country j and FDI_INjt is the value of inward of foreign direct investment of country j, Kjt is the capitalstock of country j, MANUjt is the value of manufacturing output of country j, and DISTij is the distances between Taiwan and trading partners, INDUS_DISTijkt is the trade volume of a specific industry divided by total volume of trade from Taiwan to its trading partners in year t , MANU_LABOjt is the number of manufacturing labor in country j, and the trade bloc dummies, including EUjt,, NAFTAjt, AFTAjt,, MERCOSURjt,, APECjt, which are set to considered to be one if trading partners are acting in that regional bloc in time t and otherwise zero. And this study further uses EX_lnYi_lnYj_ijkt to replace EXijkt, i.e. β1 is set to be one to form another regression.. Controlling for the year and country-industry pair effects (Model FECI) Most of the earlier studies of bilateral trade using cross-sectional models usually yield biased results due to ignoring the effect of the heterogeneity of country-industry pair 27.

(34) effects. This study also estimates the two-way fixed effect model as one of the estimations. The measure can be specified as: lnEXijkt=αt+αjk+β1lnYjt+β2lnNjt+β3lnFDI_OUTjt+β4lnFDI_INjt +β5lnKjt+β6lnMANUjt+β7lnMANU_LABOjt +β8lnINDUS_DISTijkt+β9EUjt+β10NAFTAjt+β10AFTAjt +β11MERCOSURjt+β12APECjt+εijkt .. (FECI). In the above equation, (FECI) model is controlling for the year and country-specific effects, and αjk is the interaction term of country j and industry k. We also use an alternation to replace EX_lnYi_lnYj_ijkt with EXijkt in equation (FECI), let β1 equals to one and forming the other regressions. And the other variables and dummies are noted in equation (FEIT).. Controlling for the year, country, and industry specific effects (Model FE3). Mátyás (1997) firstly introduces the three-way fixed effect of gravity specification with a dummy of time, and other two dummies of time-invariant, exporting and importing country effects to study trade flows. Mátyás (1998) further introduces that the sample with larger cross-section country specific effects should be treated as unobservable variables. lnEXijkt=αt+αj+αk +β1lnYjt+β2lnNjt+β3lnFDI_OUTjt+β4lnFDI_INjt 28.

(35) +β5lnKjt+β6lnMANUjt+β7lnMANU_LABOjt+β8lnDISTij +β9lnINDUS_DISTijkt+β10EUjt+β11NAFTAjt+β12AFTAjt +β13MERCOSURjt+β14APECjt+εijkt ,. (FE3). lnEXijkt=αt+αj+αk +β1lnYjt+β2lnNjt+β3lnFDI_OUTjt+β4lnFDI_INjt +β5lnKjt+β6lnMANUjt+β7lnMANU_LABOjt +β8lnINDUS_DISTijkt+β9EUjt+β10NAFTAjt+β11AFTAjt +β12MERCOSURjt+β13APECjt+εijkt ,. (FE3). where αj and αk are the country and industry-specific effects, respectively. And the difference of two equations is that the variable of distance between two countries.. Controlling for the year, country, industry, and country-industry pair specific effects (Model FE4) Model FEFV includes the year, county, industry and country-industry specific effects. The specification is as follow: lnEXijkt=αt+αj+αk+αjk+β1lnYjt+β2lnNjt+β3lnFDI_OUTjt+β4lnFDI_INjt +β5lnKjt+β6lnMANUjt+β7lnMANU_LABOjt +β8lnDISTijkt +β9lnINDUS_DISTijkt+β10EUjt+β11NAFTAjt +β12AFTAjt+β13MERCOSURjt+β14APECjt+εijkt ,. 29. (FEFV).

(36) where in the equation FEFV, αjk is the country-industry pair effects, and the other variables are denoted in equation FEIT.. 4.3. The modified model with two-stage estimation Stage 1. lnMANUjt=αt+αj+β1lnFDI_OUTjt+β2lnFDI_INjt+β3lnKjt+β4EUjt +β5NAFTAjt+β6AFTAjt+β7MERCOSURjt+β8APECjt+εjt ,. (FEIT). lnMANUjt=αt+αj+β1lnFDI_OUTjt+β2lnFDI_INjt+β3ln(K/L)jt+β4EUjt +β5NAFTAjt+β6AFTAjt+β7MERCOSURjt+β8APECjt+εjt ,. (FEIT). where lnMANUjt is the manufacturing output of country j, and αt is the year-specific effect common to Taiwan and other trading countries, αj is the country-specific effects, where (K/L)jt is the value of labor force divided by capital stock of country j, and the other variables and dummies are defined in equation (FECI). Stage 2. lnEXijkt=αt+αj+αk+β1lnYjt+β2lnNjt+β3lnMANUjt+β4lnFDI_INjt +β5ln(K/L)jt+β6lnDISTij+β7LANGij +β8lnINDUS_DISTijkt+β9EUjt+β10NAFTAjt+β11AFTAjt. +β12MERCOSURjt+β13APECjt+εijkt , lnEXijkt=αt+αj+αk+β1lnYjt+β2lnNjt+β3lnMANUjt+β4lnFDI_INjt 30. (FE3-1).

(37) +β5lnMANU_LABOjt+β6lnDISTij+β7LANGij +β8lnINDUS_DISTijkt+β9EUjt+β10NAFTAjt+β11AFTAjt. +β12MERCOSURjt+β13APECjt+εijkt ,. (FE3-2). where in the above equations, we adopt three-way FE models to explain the determinants of bilateral trade flows. The difference of above two equations is that in equation (FE3-1) we use capital-labor ratio (K/L)jkt as the variable, and in the equation (FE3-2) , we replace it with labor of manufacturing, MANU_LABOjt. lnEXijkt=αt+αjk+β1lnYjt+β2lnNjt+β3lnMANUjt+β4lnFDI_INjt +β5ln(K/L)jt+β6lnDISTij+β7LANGij +β8lnINDUS_DISTijkt+β9EUjt+β10NAFTAjt+β11AFTAjt. +β12MERCOSURjt+β13APECjt+εijk ,. (FECI-1). lnEXijkt=αt+αjk+β1lnYjt+β2lnNjt+β3lnMANUjt+β4lnFDI_INjt +β5lnMANU_LABOjt+β6lnDISTij+β7LANGij +β8lnINDUS_DISTijkt+β9EUjt+β10NAFTAjt+β11AFTAjt. +β12MERCOSURjt+β13APECjt+εijkt ,. (FECI-2). where in the above equations, we adopt two-way FE models. And the difference between them is the variables of capital-stock ratio, K/Ljkt, and labor of manufacturing, MANU_LABOjt. 31.

(38) lnEXijkt=αt+αj+αk+αjk+β1lnYjt+β2lnNjt+β3lnMANUjt+β4lnFDI_INjt +β5lnK/Ljt+β6lnDISTij+β7LANGij +β8lnINDUS_DISTijkt+β9EUjt+β10NAFTAjt+β11AFTAjt. +β12MERCOSURjt+β13APECjt+εijkt ,. (FEFV-1). lnEXijkt=αt+αj+αk+αjk+β1lnYjt+β2lnNjt+β3lnMANUjt+β4lnFDI_INjt +β5lnMANU_LABOjt+β6lnDISTij+β7LANGij +β8lnINDUS_DISTijkt+β9EUjt+β10NAFTAjt+β11AFTAjt. +β12MERCOSURjt+β13APECjt+εijkt .. (FEFV-2). in the above equations, (FEFV-1) and (FEFV-2) are the four-way fixed effect models, controlling for the year, importer, industry specific effects, and country-industry interaction terms. Equations (FECI-1) and (FECI-2) are controlling for the year and country-industry pair effects.. 32.

(39) Table 8. Variables and Data Sources Variables. Descriptions. Manufacturing Exports in real. Measured in millions of US. value from Taiwan to. dollars.. Sources IMF World Trade Flows, 1985-2009.. destination (EXijkt) Real Gross Domestic Product. In constant 1990 millions of US. United Nations National Accounts. (Yjt). dollars.. Main Aggregates Database 2010.. Population (Njt). Measured in one thousand people. United Nations National Accounts. per unit.. Main Aggregates Database 2010.. Output of Manufacturing. Measured in millions of US. World Development Investment. (MANUjt). dollars.. Database., 1985-2009.. Number of Manufacturing. Measured in one thousand people. World Development Investment. Labor (MANU_LABOjt). per unit.. Database, 1985-2009.. Labor Force Divided by Capital. Measured in million of US. UNCTAD Handbook of statistics,. Stock (K/Ljt). Dollars.. 1985-2010.. Foreign Direct Investment. In stock of US million dollars.. UNCTAD Handbook of statistics, 1985-2010.. (Outward & Inward) (FDI_INjt & FDI_OUTjt ) Distance (DISTij). Measured in kilometers.. The World Factbook 2010 computed distance by The Chuck Taylor Web Site.. Trade Bloc Dummy (EUjt ,. 1 if trading partner belongs to the. The official website of each trade. NAFTAjt , AFTAjt , MERCOSURjt ,. specific trade bloc in year t.. bloc.. APECjt). 33.

(40) Table 9. List of partner country of Taiwan Partner countries. Continent. Partner countries. Continent. Argentina. South America. Japan. East Asia. Australia. Oceania. S. Korea. East Asia. Austria. Central Europe. Malaysia. Southeast Asia. Belgium. West Europe. Mexico. North America. Brazil. South America. Netherlands. Northwest Europe. Canada. North America. New Zealand. Oceania. Chile. South America. Norway. North Europe. China. East Asia. The Philippines. Southeast Asia. Denmark. North Europe. Poland. East Europe. Egypt. North Africa. Portugal. Southwest Europe. Finland. North Europe. Singapore. Southeast Asia. France. West Europe. Slovakia. Central Europe. Germany. West Europe. Spain. Southwest Europe. Greece. South Europe. Sweden. North Europe. Hong Kong. East Asia. Thailand. Southeast Asia. India. South Asia. United Kingdom. West Europe. Indonesia. Southeast Asia. United States. North America. Ireland. Northwest Europe. Viet Nam. Southeast Asia. Italy. South Europe. 34.

(41) Table 10. Trade agreements and selected member countries Trade bloc. Year. Member countries. enforced EU. 1957. Belgium (1957)*, France (1957)*, Germany (1957)*, Italy (1957)*, Luxembourg (1957), Netherlands (1957)*, Denmark (1973)*, Ireland (1973)*, The United Kingdom (1973)*, Greece (1981)*, Spain (1995)*, Portugal (1986)*, Austria (1995)*, Finland (1995)*, Sweden (1995)*, Czech Republic (2004), Cyprus (2004), Estonia (2004), Latvia (2004), Lithuania (2004), Hungary (2004), Malta (2004), Poland (2004)*, Slovenia (2004), Slovakia (2004)*, Bulgaria (2007), Romania (2007).. AFTA. 2002. Brunei Darussalam (2002), Cambodia (2002), Indonesia (2002)*, Laos (2002), Malaysia (2002)*, The Philippines (2002)*, Singapore (2002)*, Thailand (2002)*, Viet Nam (2002)*.. NAFTA. 1989. Canada (1989)*, USA (1989)*, Mexico (1994)*.. APEC. 1989. Australia(1989)*, Brunei (1989), Canada (1989)*, Chile (1994)*, China (1991)*, Hong Kong (1991)*, Indonesia (1989)*, Japan (1989)*, Korea (1989)*, Mexico (1993)*, New. Zealand. (1989)*, Papua. New. Guinea. (1993),. Philippines (1989)*, Peru (1998), Russia (1998), Singapore (1989)*, Taiwan (1991)*, Thailand (1989)*, USA (1989)*, Viet Nam (1989)*.. MERCOSUR. 1994. Argentina (1994)*, Brazil (1994)*, Paraguay (1994), Uruguay (1994), Venezuela (2006).. Note: Total number of member countries of each trade bloc is shown in a parenthesis in column one. In column three the member in a parenthesis is the year when a member country enters into the trade bloc. * denotes a country that is selected member in the dataset.. 35.

(42) V. The Estimation Results This section presents regression results for different version of the gravity model. Section one introduces the models of FECI, FEIT, FE3 and FE4. In section two, we further developed the gravity model in two stages, in the first stage, we used regressions to predict the values of manufacturing output in country j, and in the second stage, we adopted the predicted manufacturing output of country j as a variable to explain the determinants of bilateral trade. In sum, this study introduces five versions of the gravity model as described in section four, which displayed and discussed the results of the estimation. In table 11, the estimation of the FEIT model is displayed in column one; he two-way fixed effect model with country-specific effects is displayed in column two and column three, and with country-industry pair effects in column four. The numbers of observations and estimation parameters, the residuals sum of squares, the adjusted R squared, the maximized value of log-likelihood, and the Akaike information criteria (AIC) and Schwarz criteria (SC) are also summarized for each model. This study judges empirical power using the adjusted R2, since this measure takes explicit account of the number of explanatory variables used in the equation. Also, the either AIC or SC is applied to serve as the decision rule for selecting a moderate model with the smallest value. 36.

(43) 5.1. The results for the FEIT and FECI gravity models In the following, we use the pooled data described in the last chapter to estimate the FEIT and two-way fixed effect model with year and country , and country-industry pair effects respectively. The regressions are reported in Table 11. The results of estimating the coefficients of the FEIT model are displayed in the first column of Table 11. The estimation indicates that bilateral exports are positively related to the importer countries’ GDP and population, and negatively related to the distance between Taiwan and its partner countries. According to the results, a 10% rise in an importer’s GDP should be associated with a 2.89% decrease in exports. Population is generally used to indicate country size. The populous countries are assumed to be a larger area endowed with a greater production and variety of natural resources, which leads to less reliance on international trade and an the expectation of a negative coefficient with the countries’ populations. From column four of Table 1, we were able to determine that if an importer’s population rises by 10%, Taiwan’s exports will decrease by 4.17%. As mentioned in Cheng and Wall (2005), and Cheng and Tsai (2008), a negative (positive) coefficient importer’s population indicates that the industry’s output is a luxury (necessity) in consumption. As shown in column one of Table11, exports decrease by 3.62% as the distance increases in increment of 10%. 37.

(44) Note that trade bloc dummies take the value of one for the country belonging to that specific bloc. This study includes five such dummy variables in the model. In column one of Table 11, the coefficients of EU are positive and statistically significant, indicating that bilateral trade in the EU member countries is more intense than in the other countries. Table 11 reports the estimation results for the augmented version of the FE model. In comparison with the FEIT model results, the results for the FE3 models predict that an increase in country j’s GDP will lead to more-than-proportional decrease in its imports, and an increase in the absolute value of the coefficient on the country’s population. The second column of Table 11 reports the estimation results for the FE3 models. The coefficients on the EU dummy from estimating the FEIT model vs. the FE3 models are quite different. As shown, with regard to the trade bloc dummies, the coefficient of the FEIT model in EU is positive and statistically significantly, i.e. their intra-bloc effects are significant. This suggests that trade pattern within the EU exhibit trade creation. Specifically, it suggests that the EU led to an increase in trade of 121.6%, and the APEC led to a decrease in trade of 95.26%. The estimated effect of the EU and APEC using the FE3 models indicates that there is something special about the relationship among their members that makes them trade less with each other than the gravity variables would predict. In column 38.

(45) one of Table 11, the estimated coefficient of foreign direct investment (FDI) inward is 0.170, while in the models of FECI and FE3 in column two and column three, the estimated coefficients are 0.268 and 0.139, respectively. According to these results, in the FEIT model, a 10% rise in the importer’s inward FDI should be associated with a 1.70% rise in imports, all else constant. However, the estimated results in the FECI and FE3 models are totally different from that in the FEIT model; the imports of country j will decline from 1.39% to 2.68% while the inward of FDI rises 10%. Moreover, as shown in Table 11,in the gravity model of the FEIT, the estimated coefficient of the importer’s capital stock -0.409, as presented in column one of Table 11. This means that the imports of country j will decrease 4.09 % while capital stock will rise 10%. According to the results of FECI, if an importer’s capital stock rises by 10%, its imports should increase about 12.53%. The FEIT model also indicates the imports will decrease about 6.74% while the manufacturing output will rise by 10%. By contrast, the estimations of the FE3 models in column three and four show that, if an importer’s manufacturing output rises by 10%, its imports should rise by 8.63%, all else constant. As shown in Table 11, this study finds that various forms of the gravity model have high explanatory power given the high values of the adjusted R2 for the individual equations ranging from 0.756 to 0.917. From the estimated results, the 39.

(46) FECI model has the highest explanatory power given the adjusted R2.. 5.2. The results for the two-way FE, three-way FE, and four-way FE models. Since the results of the above estimation are not reliable enough, we further adopt the two-stage gravity model to measure the determinants of the bilateral trade of Taiwan and its trading partner. The following are statements of description for each model: (1) in the two-way FE model: in two-way FE model, we apply for the year (αt), and the country-industry pair effects (αjk), (2) in the three-way FE model, we control not only for the year (αt) and the importer (αj), but we also added industry-specific (αjk) effects, (3) in the four-way FE model, we control for the interaction term (four-way), and, except for all specific terms, we further add country-industry pair effects into the model. In Table 14, we can discover that the industries of number five and number seven have the higher estimated coefficients than other manufacturing industries in the model of FEIT. Hence, we can conclude that the industries of five and seven account for larger sales volume in bilateral manufacturing trade flows. In addition, in the column two, the FECI model indicates that industries of number seven and number nine have higher estimated coefficients than other industries when interacting with country twenty-seven, that is, they have higher sales volume than other industries. In 40.

(47) the column three of Table 14, the estimated coefficient of interaction term for country twenty-seven and industry six has higher value than others. Therefore, we can conclude that industry six has more sales volume in country twenty-seven than other industries. As illustrated in column three of Table 14, the interaction term for country four and industry eight has the lowest value, -16.523, it presents that the trade volume of industry eight is very low in country four. Again, since the results for the one-stage gravity models are not good, we further use the two-stage gravity model to examine whether a better estimation. Table 12 and Table 13 show a comparison of the results of the two-stage gravity model. As we can see from these tables, the results of the estimations are better in the Two-stage gravity model than in the one-stage gravity model and we are quite satisfied.. 41.

(48) Table 11. Regression results for various models Model. FEIT. FECI. FE3. FE3. FEFV. lngdp_jt lnpop_jt lnfdi_out_jt lnfdi_in_jt. -0.289(0.662) 0.020(5.463) -0.465(0.236) 0.170(0.189). -0.059(0.252) -1.794(0.329)* -0.324(0.124)* 0.268(0.094)*. -0.417(0.492) 3.023(4.061) -0.434(0.175)* 0.139(0.140). -0.417(0.492) 3.023(4.061)* -0.434(0.175)* 0.139(0.140). -0.638(0.346) 3.620(2.863) -0.420(0.123)* 0.170(0.099). lnk_jt lnmanu_jt lnmanu_labo_jt lndist_j lnindus_dist_jkt EU_jt APEC_jt. 0.409(1.509) -0.674(0.680) 2.526(0.954)* -0.362(5.012) -1.169(0.023)* 12.160(4.463)* -9.526(22.010)*. 1.253(0.519)* -0.008(0.225) 0.924(0.165)*. -0.552(1.123) 0.863(0.506) 2.155(0.710)*. -0.984(0.035)* 9.755(0.716)* 12.086(0.752)*. -0.552(1.123) 0.863(0.506) 2.155(0.710)* -2.182(3.725) -0.924(0.029)* 9.866(3.326) -15.861(16.356). -0.924(0.029)* 10.053(3.318)* -16.673(17.735). -0.418(0.797) 0.661(0.357) 2.722(0.506)* -3.335(2.634) -0.967(0.041)* 16.906(2.579)* -11.851(11.539). conctant. -2.454(13.462). 10.288(5.072)*. -0.440(10.003). -20.214(26.850). 2.700(7.188). Observations Parameters Log-likelihood Adj R-squared Akaike info. crt. 1449 68 -2439.533 0.756 4949.066. 1449 67 -1606.615 0.917 3473,.229. 1449 69 -2004.821 0.865 4095.6422. 1449 68 -2004.821 0.869 4095.642. 1449 70 -1443.221 0.933 3162.442. Bay. Info. Crt Sum of sqd resides. 5133.818 2460.075. 4159.451 779.226. 4322.623 1350.095. 4322.623 1350.095. 3890.893 621.897. Note: Standard errors are in the parentheses. The estimates on year dummies, exporter dummies are omitted here. * denotes significant at the 10% level.. 42.

(49) Table 12. Regression results for FE models Model Varible. FE3-1 Ⅰ. FECI-1 Ⅱ. Ⅰ. FEFV-1 Ⅱ. Ⅰ. Ⅱ. lnk/l_jt lnfdi_in_jt lnmanu_jt lndist_j lnindus_dist_jkt EU_jt NAFTA_jt. 0.655(0.363) -0.785(3.259) 3.272(1.677). 0.717(0.360) -2.370(2.795) 10.780(2.902)*. 0.262(0.222) 0.979(0.635) -0.124(0.537). 0.206(0.215) 1.011(0.737) -0.058(0.538). 0.393(0.265) 0.010(2.378) 3.015(1.224)*. 0.456(0.262) -1.323(2.031) 10.000(2.112)*. 0.695(0.597) -3.818(3.612) -4.958(3.778) -0.903(0.029)* 19.866(11.200) 25.495(15.361). 2.623(0.787)* -16.272(4.967)* -21.800(6.434)* -0.908(0.029)* 71.253(19.922)* 97.704(27.576)*. 0.465(0.118)* -1.977(0.490)* -0.960(0.327)* -0.922(0.039)* 8.107(0.694)* 3.683(0.666)*. 0.461(0.117)* -1.896(0.463)* -0.926(0.317) -0.927(0.039)* 8.128(0.693)* 3.699(0.665)*. 0.630(0.437) -3.215(2.650) -4.187(2.780) -0.880(0.042)* 19.979(8.169)* 19.903(11.237). 2.438(0.572)* -14.878(3.611)* -19.834(4.677)* -0.892(0.041)* 67.573(14.489)* 86.921(20.056)*. APEC_jt conctant. 9.811(5.312) 85.1545(61.765). 26.866(7.525)* 377.529(111.363)*. 9.875(0.754)* 11.594(5.185)*. 9.910(0.748)* 11.971(5.198)*. 15.821(3.910)* 65.810(45.264). 31.511(5.498)* 337.044(80.983)*. Observations Parameters Log-likelihood Adj R-squared. 1479 68 -2509.099 0.862. 1479 68 -2054.176 0.863. 1479 67 -1675.945 0.913. 1479 67 -1675.698 0.913. 1479 69 -1538.063 0.927. 1479 69 -1529.582 0.928. Akaike info. crt Bay. Info. Crt Sum of sqd resides. 4200.199 4417.463 1402.052. 4190.351 4407.615 1392.748. 3609.890 4293.477 835.112. 3609.396 4292.982 834.833. 3348.126 4068.807 693.058. 3331.165 4051.845 685.155. lngdp_jt lnpop_jt. Note: Standard errors are in the parentheses. The estimates on year dummies, industries dummies, importer dummies, and dummies of interaction terms are omitted here. * denotes significance at the 10% level.. 43.

(50) Table 13. Regression results for FE models Model Varible. FE3-2 Ⅰ. FECI-2 Ⅱ. Ⅰ. FEFV-2 Ⅱ. Ⅰ. Ⅱ. lnfdi_in_jt lnmanu_jt lnmanu_labo_jt lndist_j lnindus_dist_jkt EU_jt NAFTA_jt. -0.006(0.366) 1.678(3.349) 0.239(0.396). 0.019(0.371) 0.719(2.789) -0.520(1.744). 0.238(0.218) -0.150(0.493) 0.354(0.118)*. 0.184(0.212) -0.173(0.485) 0.347(0.116)*. -0.339(0.265) 3.042(2.422) 0.296(0.287). -0.310(0.270) 1.919(2.018) 0.132(0.223). -1.364(2.304) 2.538(0.705)* 1.515(1.547) -0.913(0.029)* 13.346(3.677)* 3.584(0.548). 0.101(0.308) 2.426(0.671)* 1.858(1.506) -0.913(0.029)* 12.544(3.351) 4.243(0.667)*. -1.732(0.489)* 0.774(0.168)* -0.711(0.325)* -0.935(0.038)* 8.943(0.711)* 4.485(0.670)*. -1.653(0.462)* 0.769(0.168)* -0.673(0.314)* -0.939(0.038)* 8.953(0.710)* 4.474(0.667)*. -1.579(1.673) 3.001(0.517)* 1.983(1.161) -0.924(0.042)* 19.730(2.849)* -8.341(4.911). -0.579(1.263) 2.874(0.494)* 2.407(1.125)* -0.927(0.042)* 18.851(2.663)* -8.694(4.829). APEC_jt conctant. -3.462(5.183) -21.353(30.231). -2.075(4.426) -23.367(31.669). 10.644(0.767)* 14.295(4.477)*. 10.678(0.760)* 14.869(4.508)*. 1.940(3.778) -35.618(22.127). 3.575(3.226) -38.299(23.138). Observations Parameters Log-likelihood Adj R-squared. 1479 69 -2054.419 0.863. 1479 69 -2.54.553 0.863. 1479 67 -1664.401 0.914. 1479 67 -1664.281 0.914. 1479 69 1523.052 0.929. 1449 69 -1523.426 0.929. Akaike info. crt Bay. Info. Crt Sum of sqd resides. 4190.838 4408.102 1393.207. 4191.107 4408.371 1393 .460. 3586.801 4270.388 822.177. 3586.561 4270.148 822.043. 3318.103 4038.784 679.131. 3318.853 4039.533 679.475. lngdp_jt lnpop_jt. Note: Standard errors are in the parentheses. The estimates on year dummies, industries dummies, importer dummies, and dummies of interaction terms are omitted here. * denotes significance at the 10% level.. 44.

(51) Table 14. Regression results for various models Model. FEIT. Indus_1 Indus_2 Indus_3 Indus_4 Indus_5 Indus_6. (dropped) 1.087(0.131)* 2.844(0.144)* 2.487(0.148)* 3.118(0.154)* 2.512(0.152)*. (dropped) 0.055(0.458) 2.854(0.491)* 2.304(0.480)* -11.941(3.103)* -12.462(3.104)*. Indus_7 Indus_8 Indus_9 Coun27_ind1 Coun27_ind2 Coun27_ind3 Coun27_ind4 Coun27_ind5. 3.295(0.167)* 2.909(0.168)* 2.953(0.155)*. 1.488(0.347)* 4.994(0.423)* -0.155(0.286) (dropped) 5.716(0.722)* 7.031(0.778)* 7.190(0.798)* 21.854(3.183)*. Coun27_ind6 Coun27_ind7 Coun27_ind8 Coun27_ind9. FECI. (dropped) 5.828(0.579)* 9.969(0.611)* 9.619(0.641)* 9.961(0.594)*. FEFV. 9.693(0.601)* 22.089(3.194)* 10.008(0.590)* 8.483(0.653)* 9.631(0.594)* 4.589(0.702*) 10.013(0.591)* 10.131(0.597)*. Note: country 27 is Norway, and industries 1 to 9 are Manufacture of coke and refined petroleum products Manufacture of chemicals and chemical products Manufacture of rubber and plastics products Manufacture of basic metals Manufacture of fabricated metal products, except machinery and equipment Manufacture of computer, electronic and optical products Manufacture of electrical equipment Manufacture of machinery and equipment n.e.c. Manufacture of motor vehicles, trailers and semi-trailers, respectively.. 45.

(52) Table 15. Regression results for various models Model. FEIT. Industry_1 Industry_2 Industry_3 Industry_4 Industry_5 Industry_6. (dropped) 1.087(0.131)* 2.844(0.144)* 2.487(0.148)* 3.118(0.154)* 2.512(0.152)*. (dropped) 0.055(0.458) 2.854(0.491)* 2.304(0.480)* -11.941(3.103)* -12.462(3.104)*. Industry_7 Industry_8 Industry_9 Coun4_ind1 Coun4_ind2 Coun4_ind3 Coun4_ind4 Coun4_ind5. 3.295(0.167)* 2.909(0.168)* 2.953(0.155)*. 1.488(0.347)* 4.994(0.423)* -0.155(0.286) (dropped) (dropped) -1.142(0.392)* -13.299(3.088)* 0.623(0.419) -14.342(3.095)* -0.296(0.415)* -14.730(3.096)* 0.129(0.439)* (dropped). Coun4_ind6 Coun4_ind7 Coun4_ind8 Coun4_ind9. FECI. FEFV. -0.394(0.441)* (dropped) 1.011(0.459)* -12.521(3.103)* 0.518(0.456)* -16.523(3.132)* 0.760(0.440) -11.153(3.128)*. Note: country 4 is Austria, and industries 1 to 9 are Manufacture of coke and refined petroleum products Manufacture of chemicals and chemical products Manufacture of rubber and plastics products Manufacture of basic metals Manufacture of fabricated metal products, except machinery and equipment Manufacture of computer, electronic and optical products Manufacture of electrical equipment Manufacture of machinery and equipment n.e.c. Manufacture of motor vehicles, trailers and semi-trailers, respectively.. 46.

(53) VI. Conclusions This dissertation attempts to study the determinations of trade flows using traditional and modified models of gravity type. A pooled cross-section model is used to serve as a benchmark. A two-way fixed effect model is the main model here discussed, and Three-way and Four-way fixed effect models are estimated for comparison. The purpose is to investigate whether bilateral trade flows are affected by trading blocs, and how different distances affect bilateral trade flows. Based on the results, this study concludes that conventional variables such as the importer’s income have negative effect on bilateral trade, and that the coefficient of the importer’s population is positive.. To provide an appropriate comparison of the estimation results of conventional and modified models, this study analyzes the determination of bilateral trade flows for different importing countries with various trading blocs. As we notice, recent multilateral trade barriers under GATT and WTO are declining, and preferences among trading partners are not considered in the standard gravity models.. Based on my results, the REA effects from estimating the FEIT model vs. the FE models are not totally the same. In the FEIT model, the coefficient of variable APEC is negative, but in the FECI model, the coefficients become positive; again, in the FE3 model, they become statistically significant and positive. Moreover, in the Two-stage 47.

(54) gravity models, the estimated coefficients of EU are all positive and almost all of them are statistically significant. With the exception of the FE3-2 model, these resuts indicate that the trade patterns both within APEC and within NAFTA exhibit trade creation.. In this study, we investigate the difference in industry distance between Taiwan and other trading partners. Here we define industry distances as the total trade volume of a specific industry to a partner country divided by the total trade volume of all industries to a partner country. We find that the estimated coefficients of industry distance are almost negative. We conclude that the smaller is the volume of the product trade, the longer is the product-distance of the product is. Obviously, the variable of industry distance is worthwhile taking to taking into consideration when estimating gravity equations.. Although this study already includes more countries than pervious literature, particularly with regard to the effect of economic agreements on trade flows, it does not include all data for selected variables of all countries, most notably countries such as Indonesia, Argentina and Belgium, etc. Failure to obtain appropriate proxies for transportation costs and trade policy can cause some biases. In this study, for solving the estimation biases, I have proposed a two-way FE gravity model in which the country-specific intercepts capture the effects of trade policy and transport costs. 48.

(55) A further area of investigation would be to include the variables that measure tariff and non-tariff barriers in the estimation. Different from past studies, this study includes the output values of ten manufacturing industries’ which are the top ten in Taiwan. It further matches these with their values of exports to and imports from other trading countries. Our study results indicate that participating in a regional economic agreement does, in fact, have a significant positive effect on bilateral trade flows.. 49.

參考文獻

相關文件

The value of Textile & garment exports shrank by 73.1% year-on-year to account for 34.6% of the total exports in the first ten months of 2009; meanwhile, the value of

The value of total merchandise export rose slightly by 0.1% year-on-year to MOP694 million, with value of re-exports rising by 12.6% to MOP506 million, but that of domestic

Despite the 16.9% year-on-year leap in merchandise exports, the growth in the economy of Singapore disappointed from 3.0% over a year earlier in the fourth quarter of 2002 to only

(c) Draw the graph of as a function of and draw the secant lines whose slopes are the average velocities in part (a) and the tangent line whose slope is the instantaneous velocity

In order to achieve the core competencies of the twelve-year compulsory education curriculum and to further develop test items for science literacy assessment, it is

Valor acrescentado bruto : Receitas do jogo e dos serviços relacionados menos compras de bens e serviços para venda, menos comissões pagas menos despesas de ofertas a clientes

The value of total merchandise export for January 2011 amounted to MOP656 million, up by 5.8% year- on-year, of which value of domestic exports increased by 17.0% to MOP249 million,

substance) is matter that has distinct properties and a composition that does not vary from sample