89 Development and Challenge of the Japan-Korea-China International Input-Output Table Mitsuo Yamada Kozo Miyagawa Zaizhe Wang Abstract The economic globalization has strengthened the interdependence among countries in the world. This interdependence can be analyzed by using the international input- output (IO) tables. The Institute of Developing Economics, Japan External Trade Organization (IDE-JETRO) and the Ministry of Economics, Trade and Industry (METI) in Japan have led in the field of international IO tables for a long time. In recent years, the international IO tables have gained increasing interest from the value-added trade and value chain viewpoints. The development of the World Input-Output Database (WIOD), and Organization for Economic Co-operation and Development s (OECD) IO databases is well known. Moreover, Yokohama National University s Global IO database (YNU-GIO) in Japan has also been recently Faculty of Economics, Chukyo University, Aichi, JAPAN, yamada@mecl.chukyou.ac.jp. Faculty of Economics, Rissho University, Tokyo, JAPAN, kmiyagawa@ris.ac.jp. Faculty of Economics, Rissho University, Tokyo, JAPAN, wzz@ris.ac.jp.
90 66 1 2 made accessible to the public. The WIOD and YNU-GIO databases intended to cover relatively wide areas including more than 30 countries, although the sector classification is not so detailed. On the other hand, METI tried to build bilateral IO tables, that is 2000/2005 Japan-US table and 2007 Japan-China table. The Japan- US tables have adopted a relatively detailed sector classification with more than 170 sectors, while the Japan-China table has 77 sectors. The IDE s Asian IO table has a moderate size with 10 regions and a relatively large sector size with 76 sectors. We are conducting the study to construct the 2012 international IO table of Japan, South Korea, and China to analyze their interdependence, which is rapidly increasing through the foreign direct investment and international commodity trade. After reviewing the recent experiences of international IO databases, we discuss our IO table framework, characteristics of sector classification, advantages and disadvantages compared to other international IO tables, and future analysis we expect to conduct. Keywords International input-output table, International trade structure 1. Background and Aim The economics globalization has strengthened the interdependence among countries in the world. This interdependence can be analyzed by using the international input-output (IO) tables. The Institute of Development Economics, Japan External Trade Organization (IDE-JETRO) and the Ministry of Economics, Trade and Industry (METI) in Japan have led in the field of international IO tables for a long time. In recent years, the international IO tables have gained increasing interests from the value-added trade and value chain perspectives. The develop-
Development and Challenge of the Japan-Korea-China International Input-Output Table 91 ment of the World Input-Output Database (WIOD hereafter) and the Economic Co-operation and Development s Inter-Country Input-Output Table (OECD-ICIOT hereafter) databases is well known. Fu rther, Yokohama National University s Global IO (YNU-GIO hereafter) database in Japan has also been recently accessible to the public. We conduct the study to construct the 2012 international IO table of Japan, South Korea, and China to analyze their interdependence, which is rapidly increasing through the direct investment and the international commodity trade. In this study, after reviewing the recent experiences of the international IO databases, we introduce our project of the 2012 Japan-China-Korea International Input-Output Table (IOT) in collaboration with the State Information Center (SIC) and the National Bureau of Statistics (NBS), China, and the Korea Institute for Industrial Economics and Trade (KIET), Korea. 2. Comparison of International IOTs 2.1 Experiences in Japan First, we briefly review the experiences in Japan in the field of the international IOTs. METI in Japan has contributed to the construction of International IOTs. Table 1 shows their outcome in this field. The first international IOT compiled by METI was the 1985 Japan-US International IOT (preliminary version), which was released in 1989. Such bilateral IOTs with the UK, France, West Germany, and the US were also created within a few years. In 1993, they were combined as a multi-country IOT which is the 1985 Japan-US-EC-Asia International IOT. From 1995 to 1999, the 1990 bilateral IOTs for the US, the UK, France, and West Germany were compiled, each of which was integrated as a multi-country IOT for 1990.
92 66 1 2 Table 1. The International IOTs of METI, Japan Name of Tables Sep-89 1985 Japan-US International IOT (Preliminary Version) Mar-92 1985 Japan UK International IOT Mar-92 1985 Japan-France International IOT Mar-92 1985 Japan-West Germany International IOT Mar-93 1985 Japan-US International IOT May-93 1985 Japan-US-EC-Asia International IOT May-93 1985 Japan-US-UK-France-West Germany International IOT Sep-95 1990 Japan-US International IOT (Preliminery Version) Mar-97 1990 Japan-UK International IOT Mar-97 1990 Japan France International IOT Mar-97 1990 Japan-West Germany International IOT Oct-97 1990 Japan-US International IOT Jan-99 1990 Japan-US-EU-Asia International IOT Jan-99 1990 Japan-US-UK-France-West Germany International IOT Oct-99 1995 Japan-US International IOT(Preliminary Version) Oct-00 1995 Japan-US International IOT May-05 2000 Japan-US International IOT May-13 2005 Japa-US International IOT Mar-12 2007 Japan-China International IOT Source: METI website. Type of Tables Bilateral Tables Multi-lateral Table Bilateral Tables Multi-lateral Table Bilateral Tables After 2000, they concentrated their efforts on the IOT between Japan and the US; the 1995 IOT was released in 2000, the 2000 IOT was released in 2005, and the 2005 IOT was released in 2013. Moreover, METI conducted another trial for developing the 2007 Japan-China International IOT with the NBS, China. This project started in 2009, and the table was released in 2012 1. This project was the first trial in the sense that the International 1 The workshop was held in Beijing on March 2, 2012 for the final project report and some application studies using the IOT. (http://www.meti.go.jp/statistics/ tyo/kokusio/nittyuu/2007nen/gaiyou.html). For related studies, see Wang, Miyagawa, and Yamada (2016); Yamada, Wang, and Miyagawa (2016); and Yamada (2014).
Development and Challenge of the Japan-Korea-China International Input-Output Table 93 Table 2. The international IOTs of IDE-JETRO 1970 1975 1985 1990 1995 2000 2005 Asian International IO Table BRICs International IO Table Japan-China-Korea Asian IOT Japan-China Trans-regional Asian IOT ASEAN international IO Table Japan-Korea international IO Table Japan-Philippines International IO Table Japan-Thailand International IO Table Japan-Indonesia International IO Table Japan-China International IO Table Japan-Singapore International IO Table Japan-Malaysia International IO Table Japan-Taipei International IO Table Source: Jetro-IDE website. IOT was created through the cooperation between the two countries governments. As the second experience in Japan, we could see the Asian International IOT series. Here, IDE-JETRO is also a great contributor in this field. Table 2 shows their outcome. The Asian International IOTs have more than 20 year history. The latest IOT is the 2005 Asian International IOT in which 10 countries and regions, namely Japan, the US, China, South Korea, Chinese Taipei, Singapore, Malaysia, the Philippines, Thailand, and Indonesia are included, and the transactions among 76 sectors for each country are captured. However, no IOT after 2005 has been released to the public. 2.2 Recent Trend in International IOTs In recent years, the international IOTs have gained increasing interest from the value-added trade and value chain perspectives. The development of the WIOD and the OECD IO databases is well known. Further, YNU-
94 66 1 2 Table 3. Typical International IOT databases, easy to access Name of Tables Institutions Sectors Counties and Regions Years Asia International IOT IDE-JETRO 76, 78 10 1990/1995/2000 /2005 YNU-GIO World inputoutput database (WIOD) Yokohama National Univ. Dietzenbacher and et.al. 35 27 1997 2012 35 40 1995 2011 Inter-Country input output table (ICIOT) OECD 34 62 1995, 2000, 2005, 2008 2011 GIO database in Japan has also been recently open to the public access. Table 3 shows the typical international IOT databases, which are relatively easy to access. Besides the IDE-JETRO s Asian IOT we select three databases: the WIOD (Dietzenbacher et al. (2013), the OECD-ICIOT (Yamano (2016), and the YNU-GIO (Sato and Shrestha (2014). The WIOD, OECD-ICIOT, and YNU-GIO databases intended to cover relatively wide areas including more than 30 countries, although the sector classification is not so detailed. As another characteristic, the time series data are estimated, which enables to capture the trend and change of the international transaction structure. Figure 1 shows the characteristic of the International IOTs from the viewpoint of the area coverage and number of sectors in each table. The WIOD, ICIOT-OECD, and YNU-GIO databases intended to cover relatively wide areas including more than 30 countries, although the sector classification is not so detailed. On the other hand, METI tried to build bilateral IO
Development and Challenge of the Japan-Korea-China International Input-Output Table 95 tables, that is, 2000/2005 Japan-US table and the 2007 Japan-China table. The Japan-US tables have adopted relatively detailed sector classification with more than 170 sectors, while the Japan-China table has 77 sectors. The IDE s Asian IOT has moderate size with 10 regions, and relatively large sector size comprising 76 sectors. To build international IOTs, it has become a trend to include more countries and regions which enable to investigate the international linkage among countries, for example, in terms of trade in value added (TiVA). The more countries and regions included in the IOT, the less the number of defined sectors, because the IOT of each country has different sector definitions, and the common classification in the international IOT tends to be incomplete. This tendency makes the analysis, for example, intra-industry trade analysis of the electrical machinery sector, inadequate. There might be another direction to extend international IOTs, in which the IOT is compiled with relatively large sector size although the coverage Our Target Area Figure 1. No. of Sectors 174 Comparison of the International IOTs J-US-IO 77 76 J-C-IO IDE-AIO 35 34 YNU-GIO WIOD ICIOT 2 10 27 40 62 No. of Na ons and Regions
96 66 1 2 of regions is relatively limited. The international IOT that we intend to compile in our project belongs to this target area. A relatively large sector size allows us to investigate the industrial and trade structures in detail, impacts of lower tariffs in the framework of free trade agreements (FTAs), integrational price difference, and the industrial productivity. 3. Preliminary Studies for Our Project Here, we show some activities, which were begun prior to our project. 3.1 METI s 2012 Japan-China International IOT METI of Japan and NBS of China cooperatively constructed the 2007 Japan-China International IOT. The project started in 2009 and the table was published in June 2012. This was the first collaborative project, so both the organizations intended to build a more complete IOT in the subsequent trial including some unresolved issues. METI conducted preliminary studies for the construction of the 2012 Japan-China IOT for three years from 2012 to 2014, during which we were able to join. The issues examined were as follows: Remaining problems in constructing the 2007 IOT; Common classification of sectors; Discrepancy of international trade data between two countries; and System of national accounts (SNA) related issues and others 3.2 Issues discussed in Preliminary Studies, 2012 2014 1) Common classification of sectors There were 77 sectors in the 2007 Japan-China International IOT, and we recognized that it was necessary to increase the sector size of the 2012 Japan-China IOT. The number of sectors of 2012 China IOT is 135, and
Development and Challenge of the Japan-Korea-China International Input-Output Table 97 that of 2012 Japanese updated IOT is 516 395. We compared the classifications of 2012 IOTs in both countries, and examined the possibility of increasing the number of sectors up to around 90. It was possible to divide the following five sectors; 26 other chemical products, 35 nonmetallic mineral products, 36 Steel, 37 non-steel metals, 39 general purpose and special purpose machinery and 40 industrial electric equipment. 2) Creation of the 2012 HS-IO converter for Japan and China. To compile the international trade flow matrix by sector between Japan and China, we had to make a HS-IO converter table for each country, because the international trade statistics are compiled under the Harmonized System (HS) code, which is defined by the World Customs Organization (WCO). Although only the first six digits of the HS code are common to all countries, the following digits are used depending on the situation of each country. In the case of Japan, three digits follow making the code nine digits. China has 10 digits for its HS code. Therefore, comparing the trade between Japan and China, the HS code does not completely give a one-to-one correspondence for each commodity. However, the trade data compiled under this code system give one criterion to evaluate the commonness of the sector classification in the international IOT. The HS-IO converters for each country are important because they are used in defining the common sector classification in the international IOT. 3) International freight and insurance between Japan and China The estimation of the international freight and insurance between Japan and China is necessary to distinguish the trade transaction in producer price from that of the cost, insurance, and freight (CIF) base. In compiling the 2007 Japan-China International IOT, the cost of the international freight
98 66 1 2 and insurance was estimated by using some sampling data of Japanese international transportation companies. Therefore, we have to develop method that is more appropriate for estimation. 4) Waste treatment and by-products in each country s IOT The waste treatment and by-products might differ in each country s IOT, so we must confirm the situation and adjust the country s IOTs if there were some differences in them. 5) Re-export and re-import data in international trade statistics Re-export and re-import data in the international trade statistics cause the trade discrepancies between two countries. China s trade includes the re-export and re-import through Hong Kong in large portion. Moreover, it includes re-import of Chinese products from Hong Kong to China. These are confusing factors when comparing the export and import between Japan and China. Therefore, we have to consider a way to exclude them from the international trade of IOT for each country. 6) Treatment of the processing trade The re-export and re-import in China seem to have strong relations to the processing trade system, in which a manufacturing firm accepts customer s materials abroad for processing. In Chinese IOT, only the processing fees are included as product, while the materials costs are not included as the intermediate goods. However, in developing the international IOT, such intermediate goods should be estimated as actual international transactions. In the 2007 Japan-China IOT, such intermediate transactions were estimated. Hence, for 2012 IOT, we should address this issue more appropriately.
Development and Challenge of the Japan-Korea-China International Input-Output Table 99 7) Other important issues There are several issues to examine such as service trade, imputed housing rent, the definition of the real estate, and tourism industries. The estimation of trade in services from the data of the international balance of payments is different in each country s IOT, which we have to declare. In addition, the SNA of each country has a difference in terms of the definition of the construction activities abroad and patent fees. Therefore, we have to examine the definition and scope of trade in service for Japan-China IOT. Assuming that the same service as the rented house is provide, imputed rent of owner-occupied housing in Japan s IOT is estimated as the evaluated value using the general market price (gross rent). However, in China s IOT, it is calculated as a cost stacked (repair and maintenance costs, administrative expenses, and depreciation). Because there is such a difference in the domestic production of the imputed rent, it is important to address or reconcile the difference. In China s IOT, the activity of the real estate sector is captured as fixed capital formation. However, Japan s IOT records it separately into two parts: the real estate brokerage commission (real estate ownership relocation expenses) in the real estate agencies and managers sector, and the cost for land development in the construction sector. Therefore, this difference has to be considered in the Japan-China IOT. In the 2007 Japan IOT, tourism is recorded as the household consumption expenditure on the corresponding sectors such as restaurants, hotels, transportation etc. However, in the 2007 China IOT tourism expenditure of households in the tourism sector is considered bulk, and the expenditure on restaurants and hotels, eventually borne by household is credited to the intermediate inputs of that sector. Consequently, the domestic production of China s tourism sector becomes excessive compared to Japan. Therefore,
100 66 1 2 all the tourism expenses except tourism business costs are excluded from that sector, and modified to be accounted as the consumption expenses in the corresponding sector in the Japanese IOT. In the 2012 China IOT, the tourism industry is included in other business services and does not appear as an independent sector. Therefore, in the 2012 Japan-China IOT, we must consider the treatment of this tourism industry in China. If this sector is to be captured independently, the domestic production of the tourism industry should be estimated. 3.3 Three-Country Framework In 2015, however, METI suddenly decided not to continue with the construction of the 2012 Japan-China IOT mainly because of its lack of human resources. Simultaneously, METI also decided not to construct the regional Japanese IOT for the same reason. That was an unexpected decision for us. After the METI s decision, we decided to continue with the 2012 Japan- China International IOT project from the academic base, in collaborators with SIC and NBS of China. KIET of South Korea intended to build a similar international IOT for South Korea and was looking for collaborators. Therefore, we extended our plan to the 2012 Japan-China-South Korea IOT with the additional cooperation of KIET. Our new plan for 2016 2017 has been begun. 3.4 Compiling Process of International IOT The typical compiling processes of international IOT are as follows: 1) Defining the common sector classification for the international IOT to be compiled; 2) Adjusting each country s IOT to suit the definition of the common classification; 3) Estimating the import tables of ordinary trade and others;
Development and Challenge of the Japan-Korea-China International Input-Output Table 101 4) Estimating the cost of international freight and insurance; 5) Compiling the trade flow data among Japan, China, Korea, and the ROW by sector from the international trade database; 6) Compiling the import matrix, intermediate, and final products in the form of international IOT, and converting the price valuation from CIF to producer s price; 7) Estimating the export vector to ROW; 8) Estimating the tariff vector; and 9) Reconciling the sum-value of each row and column with the total product. The target is 2012 Japan-China-South Korea International IOT, which is built by using the 2012 IOT for each country. Table 4 shows the IOTs of each country. For Japan, there is 516 395 updated IOT (producer price, base year 2011). China has a 135 135 benchmark table (producer price, base year 2012) and South Korea hasestimated 161 161 updated IOT (basic price, base year 2010). Table 4. The IOT of each country IO Table Sector size Base year Price valuation Japan Updated IOT 516 395 2011 Producer price China Benchmark IOT 135 135 2012 Producer price South Korea Updated IOT 161 161 2010 Basic Price Source: Japan: METI website, 2012 updated input-output table (in Japanese) http://www.meti.go.jp/statistics/tyo/entyoio/result.html#menu01 China: NSBC, Manual of 2012 Input-Output Table (in Chinese) Korea: BOK website, 2012 updated input-output table http://www.bok.or.kr/contents/total/eng/boardview.action?menu NaviId=634&boardBean.brdid=14024&boardBean.menuid=634
102 66 1 2 We expect that the common sector classification would have 70 150 sectors. For Japan-China, 70 90 sectors might be possible, whereas for Japan- Korea, a more detailed division of 100 150 sectors is expected. Moreover, we must discuss the price valuation issue, because Japan and China s IOTs using producer price valuation, while South Korea s IOT is based on basic price valuation. 4. Future Plan and Remaining Issues The ongoing research is as follows. In December 2016, we will construct the HS-IO converter, obtain the IO-ISIC converter, and fix the common IOT classification. Then, we will compile the trade matrix database among the three countries, and accordingly analyze the trade structure. In 2017, we will construct the 2012 Japan-China-Korea IOT, and analyze the interdependency of the economic and trade structure among the three countries. In addition, we will analyze the effects of an economic partnership among them. We expect to perform the following analyses. According to the common sector classification, we will create a trade matrix focusing on Japan, China and South Korea. Using this trade matrix, we will examine and analyze the trade structure details of these three countries. Next, we will compare the published 2012 input-output tables of Japan, China, and South Korea. Then, after classifying the common sector, we will perform a comparative analysis of the national industrial structure. Connecting the trade matrix and international industry input-output tables, we will compile the 2012 Korea-Japan intermediate IOT. We will also analyze the route and size of the impact of international trade among them on each domestic economy. Finally, we will empirically analyze the effect of an FTA using the international IOT. The characteristics of the project we started are that, firstly, it involves
Development and Challenge of the Japan-Korea-China International Input-Output Table 103 three countries cooperative activity, and secondly, it is academically based. The goal of the project is to construct the 2012 Japan-China-Korea International IOT with more detailed sectors. This international IOT has a comparative advantage in that it is compiled after examining the definition of common sectors and international trade structures. Thus, it enables detailed sector analyses, although the regional coverage is limited. Acknowledgement This paper was presented at the 24 th IIOA Conference held in Seoul, Korea, July4 8, 2016. The authors are grateful for the comments and suggestion of the participants. This work was supported by Rissho University RPRA Center. Reference Dietzenbacher, E., B. Los, R. Stehrer, M. Timmer & G. de Vries (2013) The Construction of World Input Output Tables the WIOD Project, Economic Systems Research, 25(1), 71 98. Meng, B., Y. Zhang & S. Inomata (2013) Compilation and Applications of IDE- JETRO s International Input Output Tables, Economic Systems Research, 25(1), 122 142. Sato, K. and N. Shrestha (2014) Global and Regional Shock Transmission An Asian Perspective, CESSA WP 2014 04. Timmer, M. P., E. Dietzenbacher, B. Los, R. Stehrer, and G. J. de Vries (2015) An Illustrated User s Guide to the World Input-Output Database: The Case of Global Automotive Production, Review of International Economics, 23(3), 575 605. Wang, Z., K. Miyagawa, and M. Yamada (2016) Economic Analysis of the Japan- China Linkage Structure, Keiso Shobo (in Japanese). Yamada, M. (2014) Input-Output Analysis of the Interdependence between Japan and China through Japanese Overseas Production, Journal of Economic Struc-
104 66 1 2 tures, 3(3). 1 28. Yamada, M., Z. Wang, and K. Miyagawa (2016) Economic Interrelationship between Japan and the Chinese Coastal Area -An Empirical Analysis Using International and Regional Input-Output Model, Chukyo University, Institute of Economics, Discussion Paper No. 1504. Yamano, N. (2016) OECD Inter-Country Input-Output Model and Policy Implications, in Y. Xing (ed.) Uncovering Value Added in Trade: New Approaches to Analyzing Global Value Chains, World Scientific Publishing Co., pp. 47 59.