China-U.S. Trade Balance from the National Income Perspective

Li Xinru (李鑫茹)1, Chen Xikang (陈锡康)2, Duan Yuwan (段玉婉) 3 and Zhu Kunfu (祝坤福)4

1，2 Academy of Mathematics and Systems Sciences(AMSS), Chinese Academy of Sciences (CAS); University of Chinese Academy ofSciences (UCAS); Key Laboratory of Management, Decision and InformationSystems, CAS

3School of International Trade and Economics,Central University of Finance and Economics (CUFE)

4Research Institute for Global Value Chains (RIGVC),University of International Business and Economics (UIBE)

Abstract: This paper calculates the China-U.S. trade balancefrom the national income perspective based on an input-output model that differentiatesdomestic and foreign-invested companies. The result shows that due to differentdegrees of dependence of both countries on foreign production factors such asforeign capital for the manufacturing of export goods, only 87.7% of thedomestic value-added created by China’s exports to the U.S. in 2012 was China’snational income, whereas 96.2% of value-added in U.S. exports to China was U.S.national income. In the comparison of total export volume and exportvalue-added, the home country’s national income created by exports can morerealistically reflect a country’s gains from trade. In 2012, China’s tradesurplus with the U.S. stood at 102.8 billion US dollars in national incometerms, which is 61% and 22% smaller than the results in gross and value-addedterms, respectively. The implication is that the traditional trade balanceaccounting method seriously exaggerates the China-U.S. trade imbalance.

Key words: national income, China-U.S. trade balance,input-output model, foreign direct investment

JELClassification Codes: C67, D21, F14, F21

DOI:10.19602/j .chinaeconomist.2019.5.06

1.Introduction
Rapid growth in China-U.S. trade since the 2000shas been accompanied by an increasing bilateral trade imbalance. According todata from the U.S. Department of Commerce, the U.S. trade deficit with Chinaamounted to 347 billion US dollars in 2016, which represented 47% of US totaltrade deficit. Tremendous trade deficit is a detriment to bilateral economic,trade and political relations, giving rise to frequent trade frictions betweenboth countries - such frictions may escalate under the current U.S.administration that sees trade deficits as a high priority.
In the context of economic globalization, tradebalance based on total trade volume cannot reflect a real picture of valuedistribution. Under the global value chain theory, academics called forcorrecting trade balance data with value-added in exports (Johnson, 2014;Johnson and Noguera, 2017; Koopman et al., 2008, 2014; Wang and Sheng, 2014).Value-added in exports is concerned with gross domestic product (GDP), whichincludes return to foreign production factors that does not go into a homecountry’s income. Based on the balance of payments, Wang and Xing (2007) estimatethat the share of FDI profits re-invested in China dropped from 99% before 1996to around 60% in 2004. By the end of 2006, the re-investments of FDI profitscreated more than 220 billion US dollars of funds deposited in China. It shouldbe noted, however, that such profits could be remitted overseas anytime as longas such remittance meets national policies. With the rising cost of labor, someforeign investors in the manufacturing sector are considering leaving China.Potential capital flight may present significant economic losses and financialrisks to host countries, especially emerging market economies including China(Cai and Han, 2012; Han and Zhang, 2015).
Increasing openness of economies andinternationalization of capital markets give rise to the frequent flow ofcapital across borders. According to data released by UNCTAD, the annual growthrates of world FDI outflow and inflow reached 7% and 9%, respectively, during1990-2015. As important FDI destinations, China and the U.S. accounted for 7.7%and 17.3% of global FDI inflows, respectively, over the 1990-2015 period on anannual average basis. To some extent, FDI inflows boosted economic developmentand export growth in both countries (Feng et al., 2013; Lall and Narula, 2006; Panand Li, 2002; Zhang, 2005). As Figure 1 shows, foreign-invested companies have accountedfor over 50% of China’s merchandise exports since its WTO entry in 2001. Thisratio still stayed above 40% in recent years. Foreign-invested companies havealso made up over 20% of U.S. merchandise exports after 2010.



In addition, China and the U.S. show different dependenceon net factor inflows. It can be thus deduced that China-U.S. trade balances innational income termsare rather different from the results in value-added terms.Since 1981, China’s GDP-GNP difference has been widening. Zheng and Liu (2004)note that China’s overseas net income of primary distribution, particularly netreturn on investment, has been negative over the years. In other words, China’soverseas investment is in serious mismatch with its attraction of foreigncapital, and foreign-invested companies contributed much less to China’s GNPthan to its GDP. In comparison, the U.S. has a much smaller difference betweenGDP and GNP.
Giventhe huge foreign capital inflows and significant shares of exports byforeign-invested companies for both countries, this paper attempts to calculatethe China-U.S. trade balance from the national income perspective. This studyintends to correct distortions in China-U.S. trade equilibrium under thetraditional calculation approach. Reflecting a real picture of China-U.S. tradeis essential to easing bilateral trade tensions under the Trump administrationand facilitating bilateral trade negotiations and policymaking. As a brand-newperspective on trade and global value chain, the national income approach alsoserves as an important supplement to the value-added approach.
The rest of this paper is arranged as follows: Part 2 is a survey ofliterature studies. Part 3 introduces the model, estimation methodology and dataprocessing. Part 4 offers the result of the China-U.S. trade balance estimationin national income terms, and examines the China-U.S. trade balance equilibriumfor different types of trade and sectors. Part 5 is concluding remarks. 2. LiteratureReview
Trade imbalance lies at the heart of trade disputesand upsets bilateral and even global economic and trade relations. It hasalways received a great deal of attention in the field of economic research. Asthe world’s two largest economies, how China and the U.S. approach their tradeimbalance issues is of great interest to academics. Existing studies on thistopic generally fall into two categories: (a) calculations based on total tradevalue, (b) calculations based on value-added in trade.
Calculation of the trade balance based on totaltrade volume is the most common methodology for evaluating trade equilibrium. However,significant discrepancy exists between China’s and the U.S.’ official data onthe magnitudes of China-U.S. trade balance. Based on data from the U.S.International Trade Commission, China’s merchandise trade surplus with the U.S.in 2012 was 43.9% higher than the result of calculation based on China’scustoms data. Current studies mainly ascribe the discrepancytoshipping andinsurance costs, re-exportsthrough Hong Kong and mark-ups on re-exports, andoffer methods for correction. For instance, considering the FOB and CIFdifference, Feng and Liu (1999) and Yang (2008) take into account re-exports throughHong Kong and mark-ups on re-exports to correct the China-U.S. trade deficit.As China-U.S. trade in services developed rapidly after China’s WTO entry, theU.S. service trade surplus with China can offset the U.S. merchandise tradedeficit with China, to some extent. Xie and Li (2000) offer an empiricalanalysis on complementarity between trade in goods and trade in services.Taking into account trade in services, Barattieri (2014) re-evaluates the China-U.S. trade imbalance. Based on theaforesaid factors, Shen (2005) and Fung (2006) provide a methodology forsystematically correcting the trade balance by statistical adjustment.
Another perspective for accounting the bilateraltrade balance is value-added in trade. With deepening global division of labor,a country needs to import intermediate inputs from other countries formanufacturing its own export goods. The value-added in trade approach separatesthe value of imported intermediate inputs from the total trade value. Thisapproach is widely recognized by international institutions and academia as anew trade statistical standard. The most common method for calculatingvalue-added in trade is the input-output technology, which has been examined bymany theoretical and empirical studies.
With an input-output model, Wu and Shen (2004)calculate the contribution of China’s imports and exports to the value-addedand GDP of various sectors. Given the large share of China’s processing trade andits limited domestic value-added rate, Chen et al. (2012) and Zhu et al. (2013)re-evaluate the contribution of trade to China’s economic growth with non-competitiveinput-output tables that distinguish the processing trade from non-processingtrade.
Aside from studies and applications withsingle-country input-output tables in China’s case, global value chain studies basedon multi-regional input-output tables are carried out extensively over recentyears. Johnson (2014) and Johnson and Noguera (2017) create a methodology forcalculating value-added exports and evaluate the ratio of value-added export tothe value of gross exports for various countries, regions and sectors. Koopmanet al. (2010) and Koopman et al. (2014) put forward an aggregate exportdecomposition theory and evaluate the revealed comparative advantages ofvarious sectors of countries, offering a brand-new interpretation of globaltrade.
Based on the aforesaid methodology, Liu et al.(2007) and Lau et al. (2017) introduce the value-added in trade in calculating theChina-U.S. trade balance to correct the overestimated trade imbalance. In 2015,the U.S. trade deficit with China in value-added terms reached 132.7 billion USdollars, which was far below the 261.4 billion US dollars and 367.4 billion USdollars estimated by China and the U.S., respectively, in gross terms. Wang andSheng (2014) measure the China-U.S. merchandise trade balance in value-addedterms over the 1995-2009 period, which verifies the exaggeration of the China-U.S.trade imbalance in gross terms.
Nevertheless, merely replacing total trade valuewith value-added in trade is not enough. Since reform and opening-up in 1978,China has attracted huge foreign investments with its broad market and policypreferences. Foreign companies have accounted for more than 40% of China’stotal exports. While boosting China’s exports, FDI has also reaped return oncapital from the operating profits of foreign-invested companies. Domesticvalue-added induced by the exports of foreign-invested companies did not gointo China’s national income in its entirety. Li and Wang (2015) note that FDIfrom multinational companies leads to overestimated trade benefits. Duan et al.(2012) and Duan et al. (2013) provide China’s input-output tables withdifferentiation of domestic and foreign capital. Their estimation results showthat only 85.6% of China’s domestic value-added per 1,000 US dollars exports becameits national income. The same is true for the U.S., the largest FDI destinationin the world. In 2012, foreign-invested companies contributed 18% of total U.S.exports. Domestic value-added induced by U.S. exports is not equivalent to USnational income as well. As mentioned before, value-added in trade created byFDI does not entirely belong to the host country. Trade benefits measured byvalue-added can be threatened by a potential capital flight. In comparison, nationalincome induced by exports can reflect the actual control of export benefits. Thus,it can be used to measure a country’s actual gains from trade and internationaltrade benefits distribution. The fact that China’s GDP-GNP difference is highercompared with the U.S. means that China is more dependent on net factor inflow tosustain its economic growth. Presumably, trade benefits for China and the U.S.calculated from the national income perspective should be different from theresult from the value-added persperctive.In sum, this paper puts forward an innovative approachto calculate the China-U.S. trade balance, which serves as a reference for bilateralpolicymaking on trade development. 3. ModelDescription and Data Processing 3.1 Input-Output Model of China Reflecting the CountryAttribute of Primary Inputs
In 2012, China’s processing exports made up 42% ofits total exports, and exports by foreign-invested companies accounted for 50%of its total exports. Duan et al. (2013) formulate non-competitive input-outputtables that distinguishprocessing trade and foreign-invested companies to calculatecontributions of domestic and foreign-invested companies to China’s economy.This paper further separates the value-added in the tables into three parts, namely,China’s national income, U.S. national income and national income of othercountries.
We assume that an economy has n production sectors. In the non-competitive input-output tablesthat reflect the country attribute of primary inputs and distinguish processingtrade and foreign-invested companies (Table 1), the economy is divided intofour parts: production for domestic use or non-processing trade bydomestic-invested companies (DN), productionfor domestic use or non-processing trade by foreign-invested companies (FN), production for processing trade bydomestic-invested companies (DP) and productionfor processing trade by foreign-invested companies (FP). Meanwhile, we make the differentiation between domestic andimported intermediate inputs (M).Subscripts C, A and R in Table 1 denotethe country attribute of national income for China, the U.S. and othercountries respectively.

















from China’s customs, and trade data in servicesare from China’s State Administration of Foreign Exchange (SAFE). We also adoptbilateral trade data from U.S. sources, including trade data in goods from theU.S. International Trade Commission (U.S.ITC), as well as trade data inservices released by the BEA. To reveal China-U.S. trade balance in acomprehensive and objective manner, this paper also takes into account bilateralre-exports through Hong Kong and mark-ups on re-exports, which are provided bythe Census and Statistics Department of Hong Kong SAR government.
To decompose the value-added vector in the input-outputtables into the national incomes owned by China, the U.S. and other countries,we also need microdata officially released by China and the U.S., as well as bilateralFDI data released by UNCTAD. Value-added consists of four parts: net productiontax, compensation of employees, fixed assets depreciation and operatingsurplus. Among them, net production tax is the home country’s national incomeand does not need to be decomposed. Compensation of employees and return oncapital should be divided into the national income owned by home country andforeign countries according to the country attribute of employees and capital. Then,based on FDI stock data by origin released by UNCTAD, we can further decomposeforeign national income by particular country or economic entity. Specifically,by 2011, 5.49% of China’s FDI stock was from the U.S., and 0.15% of the U.S.FDI stock was from China. In this paper, we consider that 5.49% of foreign countries’national income induced by China’s exports is owned by the U.S., and that about0.15% of foreign countries’ national income induced by U.S. exports is owned byChina. 4. Estimationof China-U.S. Trade Balance 4.1 China-U.S. Trade Balance from the NationalIncome Perspective
To reflect the full picture of the China-U.S. tradebalance, this paper takes into account China-U.S. direct trade in goods, tradein services and re-exportsthrough Hong Kong, and calculates the China-U.S.trade balance in 2012 in gross, value-added and national income terms withresults shown in Table 3. From the national income perspective, China’s tradesurplus with the U.S. amounted to 102.8 billion US dollars, which was smallerby 60.6% and 21.8% compared with the surpluses estimated from the total tradevalue and value-added perspective, respectively. Such differences show that thetwo traditional trade balance calculation methods will exaggerate the China-U.S.trade imbalance by different degrees.
Table 3 displays China-U.S. trade balance by trade type.China’s direct trade in goods and trader-exports through Hong Kong with theU.S. were in surplus, while trade in services with the U.S. was in deficit.Thehuge China-U.S. surplus mainly came from direct trade in goods. Comparing thethree accounting perspectives, it can be found that from the national income perspective,China-U.S. trade surplus in goods and re-exports through Hong Kong were 58.0%and 57.0% smaller, respectively, compared with the total trade value perspectiveand 19.2% and 19.0% smaller, respectively, compared with the value-added perspective.China-U.S. trade deficit in services in national income terms was 31.4% greaterthan that in gross terms, and 23.7% greater than that in value-added terms.Traditionalaccounting methods exaggerate China-U.S. trade imbalance in goods andre-exports, while underestimate the bilateral trade imbalance in services.











the value-added rate. The difference betweenvalue-added rate and domestic national income rate reflects the extent of dependenceon foreign primary inputs.Thus, it can be concluded that compared with U.S.exports to China, China’s exports to the U.S. are more dependent on foreignprimary inputs, mainly foreign capital in this paper.
In addition, comparing national income induced bythree types of exports, as shown in Table 4, domestic economic benefits inducedby per 1,000 US dollars of China-U.S. trade in goods and per 1,000 US dollarsof re-exports through Hong Kong were both smaller than that induced by equalamounts of trade in services. The reasons are twofold. First, China-U.S. tradein goods and re-exports relied more on imported intermediate inputs thanbilateral trade in services. Second, China-U.S. trade in goods and re-exportsare more dependent on foreign primary inputs than trade in services. 4.3 National Income in Exports by Sector
As discussed in section 4.2, there are obviousdifferences between national income induced by China’s exports to the U.S. andthat induced by U.S. exports to China. They are determined by two factors:different national income in bilateral exports by sector and different sectoralstructure of bilateral exports. We will examine each of them in this section.
Before comparing the national income rate ofexports by sector, we first compare the value-added rate in China-U.S.bilateral exports by sector.  We findthat the complete value-added rate of China’s production for domestic use or non-processingtrade was higher than that of production for processing trade, which verifiesthe argument that processing trade has a low value-added rate. For China’sproduction for domestic use or non-processing trade, the complete value-added rateof domestic-invested company was higher than that of foreign-invested company inalmost all sectors. As for China’s production for processing trade, sectors withhigher complete value-added rate of foreign-invested company than domestic-investedcompany accounted for two thirds of all sectors. Compared with the U.S., China boastedhigher complete value-added rate in such sectors as clothing and leatherproducts, agriculture, forestry, livestock and fishery, as well astransportation equipment manufacturing, while the U.S. enjoyed higher complete value-addedrate in such sectors as electronics products, leasing and commercial services,as well as IT services. This finding suggests that China and the U.S. have comparativeadvantages in resource/labor-intensive sectors and capital/technology-intensivesectors, respectively.
Then, we compare the complete foreign nationalincome rateby sector. For both domestic-invested companies and foreign-investedcompanies in China, the complete foreign national income rate of production fordomestic use or non-processing trade was higher than that of production forprocessing trade. For the given type of production in China, the complete foreignnational income rate of foreign-invested companies was significantly higherthan that of domestic-invested companies. These results verify that from thenational income perspective, the heterogeneities of processing trade andforeign-invested companies are worth discussing.
In the comparison of the complete foreign nationalincome rate of China and the U.S. by sector, it can be found that except forpetroleum and coal products, chemical products and the mining sector, foreign nationalincome rate of the U.S. was smaller than that of China for all other sectors. Especiallyfor service industries, the complete foreign national income rate of the U.S.was much less than that of China. This finding explains the large share of U.S.national income induced by trade in services in its value-added in trade (98.1%),as shown in Table 4. It also helps to understand the conclusion that China-U.S.trade deficit significantly increased in national income terms compared to theresults in gross and value-added terms.
As can be seen from the above analysis, thecomplete domestic national income rate of China is smaller than that of theU.S. for most of sectos. On such a basis, heterogeneous sectoral structure ofexports increased the differences in national incomes induced by exports of thetwo countries.. In 2012, the top three sectors in which China had the highestexports to the U.S. were electronics products, clothing and leather products,as well as wholesale and retail, while the top three sectors in which the U.S.had the highest exports to China were wholesale and retail; agriculture,forestry, livestock and fishery; as well as transportation equipmentmanufacturing.Overall, the complete national income rate of China was dwarfedby that of the U.S. due to the low national income rate of China’s top sectorsto export. The implication is that the sectoral structure of China’s export tothe U.S. needs improvement. Besides, the U.S. enforces strict export controlsagainst China. Take high-tech products for which the U.S. enjoys comparativeadvantages for instance, U.S. high-tech exports to China accounted for 16.7% ofChina’s imports of similar products in 2001, but this percentage dropped to8.2% in 2016.[1]The falling share of U.S. high-tech exports to China is incompatible with thegreat power status of the U.S. in technology and the important partnershipbetween the two countries. For this reason, the U.S. should relax itsrestrictions of high-tech exports to China to improve its export structure andaddress the trade imbalance with China. 5. ConcludingRemarks
To reveal the economic benefits distribution in bilateraltrade between China and the U.S., this paper puts forward to estimate the China-U.S.trade balance from national income perspective on the basis of statisticalcorrection. The results show that since China’s exports are more dependent onforeign production factors, especially foreign capital, compared with U.S.exports, the China-U.S. trade surplus in national income terms, 102.8 billionUS dollars in 2012, was 61% and 22% smaller than the results in gross and value-addedterms, respectively. This finding suggests that traditional calculation methodshave exaggerated China-U.S. trade imbalance by different degrees. By the typesof trade, the China-U.S. trade surplus in goods and re-exports through HongKong decreased while China-U.S. trade deficit in services increased in nationalincome terms compared to the results in gross and value-added terms. By sector,the China-U.S. trade surplus mainly came from low-end manufacturing sectors,and China-U.S. trade deficit was mainly attributable to bulk commodities andhigh-tech sectors, which reflects the strong complementarity of bilateraltrade.
During his election campaign and after takingoffice, US President Donald Trump has frequently blamed China for China-U.S.trade deficit. In March 2018, President Trump signed an announcement to raisetariffs on imported steel and aluminum products, which unveiled a new round ofescalating trade frictions. However, based on this paper’s analysis andestimation, given China’s growing FDI and profit repatriation to advancedeconomies like the U.S., the U.S. trade deficit with China measured by nationalincome is not as serious as stated by President Trump. China-U.S. tradecomplementarity and profits of US-funded firms in China should not beoverlooked. In the context of economic globalization, trade war will not onlyharm both countries’ economies but also threaten the development of othercountries and regions as well.
By the end of 2017, the U.S. adopted a tax reformact to reduce income tax for its domestic firms and the cost of repatriatingoverseas profits. This move is likely to cause U.S. investments to flee China.Moreover, China’s outbound investments surpassed its FDI inflow in 2014. Withthe rapid growth in its outbound direct investments, China is expected to gainsignificant return on investments in developing countries, especially the Beltand Road countries. Changes in the trend of capital flow will influence China’sgains from exports and bilateral trade equilibrium in the national income terms.
This study serves as a reference for studies ontrade and the global value chain. Compared with total trade value andvalue-added in trade, national income induced by trade reflects the actual economicbenefits of various economies as they participate into the globalproduction-specialization system. By formulating the world input-output tablesthat reflect the country attribute of primary inputs, we can further extend theglobal value chain to the global income chain,which will offer a new researchperspective on the study of international trade benefits distribution.

[1]Research Report on China-U.S. Economic and TradeRelations released by China's Ministry of Commerce on May25, 2017.

References:

[1] Cai, Haoyi,and Huishi Han. 2012. “Falling FDI Stability and Prevention of MacroeconomicRisks.” Studies of International Finance,No.3.

[2] Chen, Xikang,and Cuihong Yang. 2011. Input-OutputTechnique. Beijing: the Science Press.

[3] Department of National Accounts of the NationalBureau of Statistics (NBS). 2012. China’sInput and Output Data of 2012. Beijing: China Statistics Press.

[4] Duan, Yuwan,Kunfu Zhu, Xikang Chen, and Cuihong Yang. 2013. “Non-Competitive Input-OutputModel that Differentiates Domestic and Foreign Have Invested Companies andTrade Modes.” Journal of Systems Scienceand Information, No.9.

[5] Feng, Danqin,Changbiao Zhong, and Yuanzhe Huang. 2013. “Study on the Effects of ForeignCapital Inflow Speed on the Export of Domestic Companies.” the Journal of World Economy, No.12.

[6] Feng, Guodao, and Zunyi Liu. 1999.“Re-Estimation of China-U.S. Trade Balance.”International Economic Review, No.5.

[7] Ge, Ming, Suping Zhao, and Ling Lin. 2016.“Decomposition of Allocation of Gains from China-U.S. Two-Way Trade: Based onGVC Decomposition Perspective.” WorldEconomy Studies, No.2.

[8] Han, Minchun, and Lina Zhang. 2015. “EmploymentEffects of FDI Withdraw from China’s Manufacturing Sector and EffectiveCountermeasures.” the Journal ofQuantitative & Technical Economics, No.9.

[9]Li, Hongyan, and Lan Wang. 2015.“Gains from Trade from the Global Value Chain Perspective: a Summary ofResearch Progress.” Journal of InternationalTrade, No.5.

[10]Liu, Zunyi, Xikang Chen, Cuihong Yang, and L. K. Cheng. 2007.“Non-Competitive Input-Occupancy-Output Model and Its Application: Perspectiveof China-U.S. Trade Surplus.” SocialSciences in China, No.5.

[11]Pan, Wenqing, and Zinai Li. 2002. “China’s ForeignTrade and Foreign Investment Attractions in the 1990s.” the Journal of World Economy, No.5.

[12]Shen, Guobing. 2005. “Difference of TradeStatistics and China-U.S. Trade Equilibrium.” Economic Research Journal, No.6.

[13]Wang, Lan, and Bin Sheng. 2014.“China-U.S. Trade in Value-Added and Gains from Two-Way Trade under the GlobalValue Chain Perspective.” Journal ofFinance and Economics, No.9.

[14]Wang, Yajiong, and Lixia Xing. 2007. “Measuring thePotential Threat of FDI Retained Profits to China’s Economy.” Journal of Financial Research, No.8b.

[15]Wu, Zhenyu,and Lisheng Shen. 2004. “Empirical Analysis of the Contribution of China’sForeign Trade to Its GDP.” the Journal ofWorld Economy, No.2.

[16]Xie, Kang,and Zan Li. 2000. “Empirical Analysis of Complementarity between Trade in Goodsand Trade in Services: Root Cause of China-U.S. Trade Imbalance.” Journal of International Trade, No.9.

[17]Yang, Rudai.2008. “Transit Trade through Hong Kong and Impact on China-U.S. TradeEquilibrium.” Economic Science, No.2.

[18]Zhang, Yue. 2005. “U.S. Policies on the Attractionof Foreign Capital and Development of Foreign-Invested Companies.” Global Science, Technology and EconomyOutlook, No.8.

[19]Zheng, Zhiguo, and Mingzhen Liu.2004. “Structure of China’s Economic Openness in View of GNP-GDP Difference.” China Industrial Economics, No.3.[20]Zhu, Kunfu, Xikang Chen, and Cuihong Yang. 2013.“Analysis of the Domestic Value-Added in China’s Exports and Determinants.” International Economic Review, No.4.