Phân tích mô hình trọng lực trong tạo thuận lợi thương mại ở khu vực Đông Bắc Á

Nội dung text: Phân tích mô hình trọng lực trong tạo thuận lợi thương mại ở khu vực Đông Bắc Á

1. A GRAVITY MODEL ANALYSIS OF TRADE FACILITATION IN NORTHEAST ASIA PHÂN TÍCH Mễ HèNH TRỌNG LỰC TRONG TẠO THUẬN LỢI THƯƠNG MẠI Ở KHU VỰC ĐễNG BẮC Á Zhuang LIU Party School of Central Committee of C.P.C Abstract Northeast Asia is attracting increasing attention because of its great contribution to the world’s economy. Due to the lack of regional level of formal institutionalization for Northeast Asian countries’ trade cooperation, policymakers turn their attention to the trade facilitation. In this paper, an augmented gravity model is used to model the relations between trade flows and trade facilitation. The estimation with Driscoll-Kraay standard errors is used to deal with the heteroskedasticity, and autocorrelation problem. World Bank’s Logistic Performance Index (LPI) is as the measure of trade facilitation. Our results show that trade facilitation has a significantly positive impact on the trade flows. We also quantify the gains from the improvement of trade facilitation. When the LPIs of the Northeast Asian countries get the highest regional level, the exports of the countries will increase by more than 100% on average and the intra-regional exports of the countries will increase by 152.82%. In the last section, we suggest that the trade facilitation should be enhanced from different aspects such as infrastructure, customs efficiency, logistics and technology. Further, the improvement of trade facilitation in Northeast Asia should be consider as a whole, all the countries should cooperate together to make the region better. Keyword: Northeast Asia; Trade Facilitation; LPI; Gravity Model Túm tắt Khu vực Đụng Bắc Á đang ngày càng thu hỳt sự chỳ ý nhờ đúng gúp đỏng kể vào nền kinh tế thế giới. Do thiếu cỏc thể chế chớnh thức cấp độ khu vực cho hoạt động hợp tỏc thương mại giữa cỏc quốc gia Đụng Bắc Á, cỏc nhà hoạch định chớnh sỏch chuyển sự chỳ ý sang hoạt động tạo thuận lợi thương mại. Trong bài viết này, mụ hỡnh trọng lực mở rộng được vận dụng để khỏi quỏt mối quan hệ giữa dũng chảy thương mại và tạo thuận lợi thương mại. Ước tớnh với sai số chuẩn Driscoll-Kraay được sử dụng để khắc phục phương sai thay đổi và vấn đề tự tương quan. Chỉ số Năng lực quốc gia về Logistics (LPI) của Ngõn hàng thế giới được coi là cụng cụ đo lường mức độ tạo thuận lợi thương mại. Kết quả của chỳng tụi cho thấy tạo thuận lợi thương mại cú tỏc động tớch cực tới dũng thương mại. Chỳng tụi cũng lượng húa những kết quả đạt được từ việc tăng cường tạo thuận lợi thương mại. Khi chỉ số LPI của cỏc quốc gia Đụng Bắc Á đạt mức cao nhất trong khu vực, xuất khẩu của cỏc nước này trung bỡnh sẽ tăng thờm hơn 100% và xuất khẩu nội khối sẽ tăng thờm 152.82%. Trong phần cuối bài, chỳng tụi đề xuất tăng cường tạo thuận lợi thương mại từ cỏc phương diện khỏc nhau như cơ sở hạ tầng, hiệu quả hoạt động hải quan, logistics và cụng nghệ. Ngoài ra, việc tăng cường tạo thuận lợi thương mại ở khu vực Đụng Bắc Á cần được xem xột một cỏch tổng thể, tất cả cỏc quốc gia cũng cần phải hợp tỏc với nhau để tạo nờn một khu vực tốt đẹp hơn. Từ khúa: Đụng Bắc Á; Tạo thuận lợi cho thương mại; LPI; Mụ hỡnh Gravity 284
2. 1.Introduction Globalization intensifies international competition while offering extensive opportunities for development. As the integral level of formal trade barriers has decreased, the increase of the “soft” power of trade is required. Thus, the topics on trade facilitation have been brought to researchers’ tables. However, there is no generally accepted definition for trade facilitation. OECD defines the trade facilitation as “streamlining and simplifying international trade procedures in order to allow for easier flow of goods and trade at both national and international level.”104 According to WTO, trade facilitation is “the simplification and harmonization of international trade procedures” where the procedures include “the activities, practices, and formalities involved in collecting, presenting, communicating, and processing data and other information required for the movement of goods in international trade.” Generally, the term refers to the ease of moving goods across the borders (Felipe and Kumar 2010). Northeast Asia plays a critical role in Asia and even in the world. The GDP of Northeast Asian countries is more than 20% of the total GDP of the world in 2014, but the trade in the region has not been well-developed. Meanwhile, the lack of regional level of formal institutionalization for Northeast Asian countries’ trade cooperation drives policymakers to focus on more practical ways to improve the regional trade. This paper uses an augmented gravity model to examine the impact of trade facilitation on the Northeast Asian countries’ export flows and estimate the changes in trade from the improvement of trade facilitation.105 The World Bank’s Logistic Performance Index (LPI) is employed as the indicator of trade facilitation. 2. Literature Review The great contribution to the world’s economy and the accelerated process of integration make Northeast Asia attract increasing attention. Lee (2005) finds the trade concentration ratio in Northeast Asia increased from 1.09 in 1990 to 1.65 in 2004 and he estimates an FTA is likely to generate economic welfare of 30 billion USD with a CGE model. However, Kuznetsova (2013) finds that formation of trade coalition does not cause a gain of mutual trade flows in Northeast Asia with a gravity model. Some studies have indicated that without reaching any regional trade agreement, lower tariff can increase the volume of trade flows.106 Besides tariff, cost of trade is a strong factor influencing Asian regional trade. Using a gravity model, De (2004) has shown that transaction cost including port efficiency and infrastructure quality is statistically significant in explaining trade in Asia. Similarly, the improvement of trade facilitation increased intra-APEC trade of the APEC countries and (Wilson, Mann and 104 105 Due to the availability of data of North Korea, Northeast Asian countries only include China, Japan, Republic of Korea, Russia Federation and Mongolia. 106 According to Lee (2005), the weighted average tariff of China, Japan and Korea in 2004 decreased to less than 6 percent from around 20 percent in 1991, and the regional trade volume in Northeast Asia increased 479% from 56 billion USD IN1991 to 325 billion USD in 2004. 285
3. Otsuki, 2003). But there is little evidence on the impact of trade facilitation on Northeast Asian intraregional trade. Different measures of trade facilitation were used to estimate the improvement of trade from enhancing trade facilitation. The difference of cost, insurance and freight, and free on board values (Limao and Venables, 2001), trade mobility index (TMI)(Wilson, Mann and Otsuki, 2003) and time taken to export (Djankov Freund, and Pham, 2006) were chosen to capture the quality of trade facilitation in the past studies. Besides, Hertel and Mirza (2009) and Felipe and Kumar (2010) employed the World Bank’s Logistic Performance Index (LPI), like this study, to quantify the gains from improvement of trade facilitation. 3.Emperical Analysis 3.1 Model Selection Gravity models are widely used to model the relation between bilateral trade flow and trade cost. The inspiration of the original gravity model is from the Newtonian universal gravity. Since the first empirical work by Tinbergen (1962) and Pửyhửnen (1963), the gravity model has been further developed to explain trade flows by different determinants. The general formulation of the gravity models explains a trade flow Tradeij from origin i to destination j with a measure of market size of the origin and destination Gi and Gj (such as GDP, GDP per capita and population), and the distance between the trade partners dij (to capture the cost of trade). Generally, gravity models are in log-linear form: lnTradeij = β0 + β1 ln Gi + β2 ln Gj + β3 ln dij + uij (1) where β0 is constant, β1 , β2 and β3 are coefficients, uij is error term. The general gravity model builds a bridge between trade flows and trade barriers. In order to specify the impact of trade facilitation on trade, we add World Bank’s LPI and dummy variables Common Border and Common Language to construct our augmented gravity model of export from exporting economy i to importing partner j in the trading year t (t=2007, 2010, 2012) as follows: ln Expijt = β0 + β1 ln GDPPCit + β2 ln GDPPCjt + β3 li n POPit + β4 ln POPjt + β5 ln Dij （2) + β6 ln LPIit + β7 ln LPI jt + β8 ln Borderij + β9 ln Languageij + uijt with dependent variables as follows (expected sign on coefficients in the parentheses): GDPPCi(j)t : the per capita gross domestic product of economy i (j) (β1>0, β 2>0); POPi(j)t : the population of economy i (j) (β3>0, β4>0); Dij : the distance between the capitals of economies i and j (β5 0, β7>0); Borderij =1 if economies i and j share a common border, otherwise 0(β8<0); Languageij =1 if economies i and j use a common language, otherwise 0(β 9<0). 286
4. 3.2 Data Sources and Methodology Due to the availability, the Northeast Asian countries only include China, Japan, South Korea, Russia and Mongolia.107 The sample data includes the export data of the five Northeast Asian economics to other Northeast Asian partners and their other nine main trade partners are selected108 in 2007, 2010 and 2012. There are 195 observations including nine variables. The export data is from the United Nations COMTRADE databases. The data of GDP per capita, population and World Bank’s LPI are taken from World Development Indicators. The distance is calculated as the distance between the capitals of the trade partners. The distance ,Common border and common language come from CEPII. The regression with Driscoll-Kraay standard errors is used to deal with the heteroskedasticity, and autocorrelation problem in the unbalanced panel data. Panel data is common to be analyzed in economics, but too much information of cross-sectional and temporal dependencies can make the panels overstated. Therefore, studies with a regression on panel data normally adjust the standard errors of the coefficient estimates for possible dependence in the residuals. An assumption of the cross- sectional independency of the disturbances in a state- level panel model usually is not inappropriate. Furthermore, if the patterns of the dependence are raised by unobservable common factors and the factors are correlation with the explanatory variables, the panel estimators with fixed effect (random effect or pooled OLS estimation will fail because of the inefficiency of the coefficient estimators. To solve the problem, Driscoll and Kraay (1998) propose a covariance matrix estimator and the estimator generates heteroskedastic and consistent standard errors that are robust to general forms of cons-sectional (spatial) and temporal dependence. Such standard error called Driscoll- Kraay standard error. In this study, the unbalanced state- level panel data is used and there may exist some common influencing factors of trade in different countries. The estimation with Driscoll and Kraay standard error is used to ensure the estimators precise and efficient. 3.3 Empirical Results Table 1 shows the results from the estimation of equation 2. The first column of Table 4 lists the OLS estimates of equation 2 in logarithms and the result of White test. The sign of estimates satisfy our expectations. The coefficient of determination R2 is 0.855. The results are well explained by the model. The main dependent variables are statistically significant and most of the estimates of the coefficients are around 1. Besides, The result of White Test is rejecting the original hypothesis that means the existence of heteroskedasticity. 107 According to World Bank, the Northeast Asia refers to China, Japan, South Korea, North Korea Mongolia and Russian Far East and Siberia. But the economic information cannot be accessed, the North Korea is excluded in this study. 108 According to the trade statistics in 2012, the nine trade partners include Australia, Hong Kong, Germany, Malaysia, Singapore, Thailand, United States, United Kingdom, Vietnam. 287
5. The GLS robust variance estimation is used, since the sample is panel data and the heteroskdasitic problem. Column 2 in Table 1 presents the results from GLS robust variance estimation and Hausman specification test. The main results of this study are shown in column 3. The sign of the coefficients are in line with our expectation in the literature and all the dependent variables are statistically significant at 1%. This confirms that estimation with Driscoll and Kraay standard error performance better in this study. The sizes of the trade partners are main factors that positively influence trade flows. Specifically, increase in GDP per capita of the exporter by 1% brings more growth in trade flow than the importer. Similarly, The coefficient of exporter’s population is 1.17, which is larger than the coefficient of importer’s population. Since China is a major importer and exporter in this region, the large population of China means a sufficient labor supply for exporting and a big demand market for importing. The basic variable Distance negatively impacts trade flows. When distance decreases by 1%, the trade flow will increase by almost 1%. A Common border or common language also increases the trade flows between trading partners as the column 3 shown. LPIs, our most interested variables, do have our expected signs. Every 1% improvement in the LPI of exporting economy increases exports by 0.78%. A raise in the LPI of importing economy by 1% only results in an increase in imports by 0.89%. This is the evidence of the importance of trade facilitation in trade. In the following section, the gain from an improvement in trade facilitation is quantified. Table 1: The Results of the Regressions and Tests of Equation 2 Dependent Variable: Log of Export (1) (2) (3) VARIABLE Random Effect Driscoll and OLS GLS Kraay Exporting economy’s 1.31 0.92 1.31 GDP per capita (0.17) (0.22) (0.04) Exporting economy’s 0.58 0.56 0.58 GDP per capita (0.16) (0.22) (0.20) Exporting economy’s 1.17 1.16 1.17 population (0.07) (0.11) (0.02) Importing economy’s 0.75 0.74 0.75 population (0.07) (0.11) (0.03) Distance 0.97 0.90 0.97 (0.18) (0.27) (0.04) Exporting economy’s 0.78 1.12 0.78 LPI (0.28) (0.40) (0.08) 288
6. Importing economy’s 0.89 0.72 0.89 LPI (0.36) (0.50) (0.31) Common border 1.83 1.51 1.83 (0.41) (0.62) (0.26) Common Language 0.93 0.64 0.93 (0.60) (0.92) (0.20) Constant 28.39 25.34 28.39 (2.54) (3.68) (1.30) R2 0.855 0.856 White Test(Chi2) 85.62 Hausman Test(Chi2) 14.01 Standard errors in parentheses * p < 0.10, p < 0.05, p < 0.01 3.4 Simulation Design and Results The results of the gravity model proved trade facilitation enhance trade of Northeast Asian countries statistically in the last section. We design two scenarios to quantify the economic gain from improving trade facilitation (LPI). Table 2 shows the average and individual LPIs of the Northeast Asian countries. The LPIs are highly various among the countries. It is not practical to make the same one-off improvement in the trade facilitation of all the countries. Considering feasibility, scenario 1 supposes the LPIs of the Northeast Asian counties are at least the average level of the region. To test the trade potential, we design the LPIs of all the regional countries are at the highest level of the region in scenario 2. Table 2: The LPIs of Northeast Asian Countries YEAR COUNTRY 2007 2010 2012 CHINA 3.32 3.49 3.52 JAPAN 4.02 3.97 3.93 KOREA 3.52 3.64 3.70 RUSSIA 2.37 2.61 2.58 MONGOLIA 2.08 2.25 2.25 AVERAGE 3.06 3.19 3.20 In scenario 1, Only Russia and Mongolia's LPIs are lower than the regional average LPIs. As table 3 shown, if the LPIs of the three countries is brought to the regional average level, the total export of the region to the economies in the sample will increase by 6.82%. As the result of the improvement of trade facilitation, the export of a Northeast Asian county to other economies in the sample will increase at least 53.6 million USD. The country with the lowest LPIs, Mongolia, gets the highest trade improvement (122.37 %) If the LPIs of all the countries in the region reach the Japan's level (regional highest level), the exports of the countries will increase 103.76 % on average. The benefits of the 289
7. countries in the region benefit from the improvement in LPI are significant. Most counties get more than 100% increase in export due to the LPI’s improvement. We can find the great trade potential in Northeast Asia. But the scenario 2 is an ideal case, it is a long way for the counties in the region to perfect trade facilitation. Table 3: Estimated Trade Gains in the Two Scenarios SCENARIO>1 SCENARIO>2 (at>least>average>LPI) (reginal>highest>LPI) AVERAGE>CHAN AVERAGE>CHAN %>CHANGE GE>IN> GE>IN> %>CHANGE>IN> >IN> EXPORT(MILLIO EXPORT(MILLION EXPORT EXPORT EXPORTER N>USD) >USD) CHINA 5900 6.35% 113100 121.74% JAPAN 1300 1.47% 43500 49.15% KOREA 100 0.55% 20600 113.19% RUSSIA 7200 67.92% 41300 389.62% MONGOLIA 53.6 122.37% 294.2 671.69% OVERALL>L 2900 6.82% 44100 103.76% EVEL The estimated gains in intra-Northeast Asian trade from the improvements in LPI also are calculated (see Table 4). In scenario 2, intra-Northeast Asian trade increases by as much as702.42%. Besides, scenario 2 increases the trade in Intra-Northeast Asia more than the trade with rest of the world. It reveals that the trade facilitation is critical for the economic integration of Northeast Asia. Table 4: Change in Intra-Northeast Asia Exports from the Improvement of LPI in Scenario 2 %&CHANGE&IN&EXPORT TRADE&WITH&ALL&OTHER& TRADE&IN& EXPORTER ECONOMIES&IN&THE& INTRA1NORTHEAST& SAMPLE ASIA CHINA 121.74% 213.16% JAPAN 49.15% 81.50% KOREA 113.19% 130.11% RUSSIA 389.62% 500.00% MONGOLIA 671.69% 702.42% 4. Policy Suggestion As discussed, an improvement in LPI does increase trades of the Northeast-Asian countries whatever the original level of LPI is. Thus, the policymakers should learn from 290
8. the experience of the countries with better trade facilitation and take some actions to strengthen the regional corporation in Northeast Asia. First, trade facilitation should be improved from different aspects. Trade facilitation, as well as LPI, includes different components, such as infrastructures, logistics, customs efficiency, tax policy and technology. Generally, poor customs efficiency and outdated infrastructure cause a high trade cost. Thus, the weakest component should be found out and improved at the first step. Increasing budget on infrastructure, adapting international trade standards and simplifying customs processes are important. Besides, the application of the new technology like IoT(the internet of tings) is also a critical part of trade facilitation. Further, the negotiations of trade facilitation and integration should be expedited. The importance of trade facilitation is a consensus around world. However, the countries in Northeast Asia have different political system, economic development and cultures. It is necessary to discuss trade facilitation in the region as a whole. Regional economic integration and improvement of trade facilitation are reciprocal causation. This also requires us to discuss trade improvement systematically. REFERENCES Djankov S., C. Freund, and C. Pham. 2006. Trading on Time. World Bank Policy Research Working Paper 3909. The World Bank, Washington, DC. Driscoll, J. C., and A. C. Kraay. 1998. Consistent Covariance Matrix Estimation with Spatially Dependent Panel Data. Review of Economics and Statistics 80: 549-560. Felipe J. and U. Kumar. 2010. “The Role of Trade Facilitation in Central Asia: A Gravity Model. Working Paper No. 628. The Levy Economics Institute Working Paper Collection . Hertel T. and T. Mirza. 2009. "The Role of Trade Facilitation in South Asian Economic Integration." Study on Intraregional Trade and Investment in South Asia. ADB, Mandaluyong City. Lee, C. J. et al. 2005. Rationale for a China-Japan-Korea FTA and Its Impact on the Korean Economy, Korean Institute for International Economic Policy (KIEP), Seoul Limao, N. and Venables, A. J. 2001. “Infrastructure, Geographical Disadvantage, TransportCosts, and Trade”, The World Bank Economic Review 15: 451-479. Wilson, J., C. Mann and T. Otsuki. 2003. Trade Facilitation and Economic Development: Measure The Impact. World Bank Policy Research Working Paper 2988. The World Bank, Washington, DC. Kuznetsova, N. V., 2013, Economic Integration of Northeast Asia, World Applied Sciences Journal 25 (5): 768-773. 291