Abstract
According to a number of studies, Chinas bundle of exports is relatively sophisticated. This paper investigates this finding by introducing an import sophistication measure and dividing products by type into Consumption, Capital and Intermediate goods. The results show that Chinas imports are actually more sophisticated than exports and highly sophisticated in relative terms. Controlling for import sophistication, income and human capital, we find that Chinas exports are highly sophisticated for Intermediate products only. Consumption goods are not highly sophisticated, while Capital goods are less sophisticated than the average. Despite being relatively sophisticated Chinas exports are of low quality according to different measures based on unit values. Once we split the market into price segments we see that the sophistication of Chinas exports is exceptional for low-priced Intermediate and Capital products only.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
- 2.
Wang-Wei’s export dissimilarity index is a simple transformation of the Finger-Kreinin index, thus there are no actual methodological differences with Schott (2008). Notice that there are other indexes similar to a sophistication index or similarity/dissimilarity indexes, for instance in Michaely (1984), Grubel and Lloyd (1975) and others. More recently, Van Assche-Gangnes (2010) and Lall et al. (2006) proposed other sophistication measures.
- 3.
Balassa’s RCA index is given by the relative country’s export share of a given product/sector (c and p are country and product subscripts):
$$ RC{A_{{c,p}}} = \frac{{{{{{x_{{c,p}}}}} \left/ {{\sum\nolimits_p {{x_{{c,p}}}} }} \right.}}}{{{{{\sum\nolimits_c {{x_{{c,p}}}} }} \left/ {{\sum\nolimits_c {\sum\nolimits_p {{x_{{c,p}}}} } }} \right.}}} $$ - 4.
The Grubel-Lloyd index measures import–export overlap. The country aggregate index is defined as:
$$ G{L_c} = 1 - \frac{{\sum\nolimits_p {\left| {{x_{{c,p}}} - {m_{{c,p}}}} \right|} }}{{\sum\nolimits_p {{x_{{c,p}}} + {m_{{c,p}}}} }} $$ - 5.
We used the following transformation GL = log[GL/(1 − GL)], which takes values from −∞ to +∞. The logit transformation does not change the results, but it allows us to get a clearer picture and to perform better regressions. Hellvin (1996) uses the same transformation.
- 6.
The GL index is said to be downward biased for unbalanced trade because in that case it can never reach the value of 1. This happens because even though every product can show balanced trade, and therefore the maximum level of IIT, there will be at least one product where trade is unbalanced (in this case the product imbalance will exactly match the overall imbalance). However, if we consider the GL index as an overlap measure, it correctly tells us that the import–export overlap is imperfect.
- 7.
Many countries show no significant correlation between imports and exports, but this is due to the fact that trade regards relatively few countries. For many countries there are very few observations or no data at all. In 1995 the countries with which China both imported and exported more than 1,000 HS6-products simultaneously (there are about 5000 HS6 codes) numbered just 16 (in 2000 it was 24; in 2007, 33) and those few countries represented more than 73% of its overall trade (78.6% in 2000; 76.5% in 2007). If we only consider simultaneous imports–exports, then that share rose to almost 96% in 1995 (96.2% in 2000; 96.6% in 2007). Only 36 countries both imported from and exported to China more than 100 products (47 in 2000; 69 in 2007), 46 traded more than 50 products (59 in 2000; 84 in 2007), 63 more than 10 products (94 in 2000; 122 in 2007). Details can be provided by the author upon request.
- 8.
An index calculated with data on imports only would have a slightly different interpretation. In that case sophisticated products would be no more than those exported by rich countries, but those that are imported by them. Therefore, if a country imports products that are imported by rich countries its imports will appear relatively sophisticated. We do not think this is the right measure, at least for our task. Note that Cui and Syed (2007) measures import sophistication in this way and gets similar aggregate results. Roughly speaking, we believe this is due to the fact that IIT is quite high for developed countries, therefore their imports and exports are relatively similar, which implies that importing products that rich countries export tends to be the same as importing products that rich countries import.
- 9.
Figure 9.7 gives some useful hints on the dynamics. However, we must warn that the time comparison is imprecise: our sophistication measures allow a cross-country comparison, while time trends in sophistication levels can partially reflect income growth, especially for exports which tend to be less diversified than imports. This point is discussed in the next section.
- 10.
In the numerical example of Table 9.4 in Kumakura (2007), in which each country completely specializes in one product (no export overlap), this would imply that expy/income = 1 for every country independently of income. In this case, relative sophistication can only change with export structure, i.e. export overlap. Notice, however, that absolute sophistication levels are now not perfectly comparable over time since the increase in the measured sophistication partially (or completely in the case of complete specialization) reflects the increase in income: Controlling for income and adding year-fixed effects solves the problem.
- 11.
Notice that we do not even have such detail in custom HS6 trade statistics, instead, we have something like “Sound apparatus… Using magnetic, optical or semiconductor media” (851981).
- 12.
Notice that in our dataset both import and export values are expressed in f.o.b. prices, which allows a consistent comparison between them. One possible limitation is that our data are in US dollars, therefore we cannot exclude that exchange-rate dynamics are partially reflected in UVs. We can reasonably say that since our analysis is quite detailed and covers 13 years, our results are not determined by exchange rates only. See data description in Appendix A.1 for details.
- 13.
A similar analysis of exports and prices is applied to Italy at a more detailed level in Marvasi (2010).
- 14.
In this case the RCA index is understood as a relative import share, indicating that a country import is relatively concentrated in some products with respect to the world.
- 15.
We also tried dividing the market into two segments according to r < 1 or r > 1. The results are quite similar. However, the three segments version is more informative and easier to compare with Fontagné et al. (2008) methodology.
- 16.
See Table A.5 for the middle-end segment. Results from the regressions under the alternative methodology can be provided by the author upon request. We also checked whether the Xu (2010) methodology was in accordance with our results. Using a low degree of price adjustment, the results are broadly in line with the overall regressions, while with a strong price adjustment they are more in line with the high/medium-end segment regressions.
- 17.
Notice that China’s dummy coefficients can be interpreted directly as percent deviation only if small. Otherwise the approximation error cannot be ignored. For instance, for high-end Consumption goods the China 04–07 coefficient is −1.5; this means that China’s export sophistication is lower than expected by approximately 78% (e−1.5 −1).
References
Amiti M, Freund C (2010) The anatomy of China’s export growth, NBER Chapters, in: China’s growing role in world trade. National Bureau of Economic Research, pp 35–56
Balassa B (1965) Trade liberalization and revealed comparative advantage. The Manchester School of Economic and Social Studies, 33:99–123
Balassa B (1967) Trade liberalization among industrial countries. McGraw-Hill, New York
Branstetter L, Lardy N (2006) China’s embrace of globalization. NBER working paper series
Cui L, Syed MH (2007) Is China changing its stripes? The shifting structure of China’s external trade and its implications. Forthcoming IMF working paper
Dean J, Fung KC, Wang Z (2008) How vertically specialized is Chinese trade? Office of Economics working paper No. 2008–09-D, USITC, Dec 2008
Dettmer B, Erixon F, Freytag A, Legault-Tremblay PO (2009) Dynamics of structural change: the European union’s trade with china. Jena Economic research papers 2009-053, Friedrich-Schiller-University Jena, Max-Planck-Institute of Economics
Feenstra RC, Hanson GH (2005) Ownership and control in outsourcing to China: estimating the property-rights theory of the firm. Q J Econ 120(2):729–761
Fernandes A, Tang H (2010) The determinants of vertical integration in export processing: theory and evidence from China. Development working papers 289, Centro Studi Luca d’Agliano, University of Milan, revised 30 Apr 2010
Findlay R (1978) An “Austrian” model of international trade and interest rate equalization. J Polit Econ 86(6):989–1007
Finger JM, Kreinin ME (1979) A measure of export similarity and its possible use. Econ J 89:905–912
Fontagné L, Gaulier G, Zignago S (2008) Specialisation across varieties within products and north-south competition. Econ Pol, CEPR, CES, MSH, 23:51–91, 01
Fontagné-Freudenberg (1997) Intra-industry trade: methodological issues reconsidered. Working papers 1997–01, CEPII Research Center
Gaulier-Zignago (2010) BACI: international trade database at the product- level. The 1994–2007 version. Working papers 2010–23, CEPII Research Center
Gilboy G (2004) The myth behind China’s miracle. Foreign Affairs, June/August, pp 33–48
Grubel HG, Lloyd PJ (1975) Intra industry trade: the theory and measurement of internationally trade in differentiated products. Wiley, New York
Hausmann R, Hwang J, Rodrik D (2007) What you export matters. J Econ Growth 12(1):1–25
Hellvin L (1996) Countries, vertical intra-industry trade between China and OECD. OECD Development Centre working papers 114, OECD, Development Centre
Hu X, Ma Y (1999) International intra-industry trade of China. Weltwirtschaftliches Archiv 135(1):82–101
Hummels D, Ishii J, Yi KM (2001) The nature and growth of vertical specialization in world trade. J Int Econ 54(1):75–96
Kumakura M (2007) What’s so special about china’s exports? A comment. China World Econ 15(5):18–37
Lall S, Weiss J, Zhang J (2006) The sophistication of exports: a new trade measure. World Dev 34(2):222–237
Lemoine F, Unal-Kesenci D (2004) Assembly trade and technology transfer: the case of China. World Dev 32(5):829–850
Marvasi E (2010) Prezzo e Qualità: in che direzione si muove l’export italiano. Rapporto ICE 2009–2010
Michaely M (1984) Trade, income levels, and dependence. North-Holland, Amsterdam
Rodrik D (2006) What’s so special about China’s exports? China & world economy, institute of world economics and politics. Chin Acad Soc Sci 14(5):1–19
Schott PK (2008) The relative sophistication of Chinese exports. Econ Pol, CEPR, CES, MSH, 23:5–49, 01
Van Assche A, Gangnes B (2010) Electronics production upgrading: is China exceptional? Appl Econ Lett, 15:477–482
Van Assche A, Hong C, Slootmaekers V (2008) China’s international competitiveness: reassessing the evidence. LICOS discussion papers 20508
Wang Z, Wei SJ (2008) What accounts for the rising sophistication of China’s exports? NBER working papers 13771, National Bureau of Economic Research
Xing Y (2007) Foreign direct investment and China’s bilateral intra-industry trade with Japan and the US. J Asian Econ 18(4):685–700
Xu B (2010) The sophistication of exports: is China special? China Econ Rev 21:482–493
Xu B, Lu J (2009) Foreign direct investment, processing trade, and the sophistication of China’s exports. China Econ Rev 20:425–439
Yao S (2009) Why are Chinese exports not so special? China Econ Rev 17:47–65
Acknowledgments
I am grateful to Alessandro Borin, Barbara Annicchiarico and Massimo Armenise for their useful criticism and comments. I must also thank Beniamino Quintieri and Giorgia Giovannetti for starting this research project and getting me involved. Special thanks go to all the staff of the Manlio Masi Foundation for their support. All errors are my own.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Appendix
Appendix
1.1 A.1 Dataset Description
Trade data used in this paper are taken from the BACI dataset from CEPII, which is built from the UN-COMTRADE database. The dataset reports HS 6-digit bilateral trade flows of all available countries from 1995 to 2007. The main advantage of BACI is that it reconciles the declarations of the exporter and the importer, in particular both imports and exports are in f.o.b. prices, which enables a proper comparison of UVs. For a complete description of the dataset see Gaulier-Zignago (2010) “BACI: International Trade Database at the Product- level. The 1994–2007 Version” and www.cepii.fr. Data on real per capita income, real GDP, population and school enrollment are from the World Bank. Real per capita GDP and real GDP are expressed in dollars at 2,000 constant prices. Secondary school enrollment (gross) is the ratio of total enrollment, regardless of age, to the population of the age group that officially corresponds to the level of education shown. Data on the import content of exports are taken from the OECD.
1.2 A.2 Market Price Segments
The methodology we adopted to split the market into price segments is taken from Fontagné et al. (2008). The idea behind it is that there is a continuum of vertically differentiated varieties. Product UV is therefore an average of the prices of these varieties. A product with a relative UV (r) that is just above one must include within it a lower share of high-priced varieties than a product with much greater UV. Following this idea each single trade flow is split into price segments according to the following scheme: When r ≤ 1, the low-end share is (1 − r α) and the middle-end share is (r α); when r > 1, the middle-end share is (1/r α) and the high-end share is (1 − 1/r α).
Following Fontagné et al. (2008), who use the same data, we set α = 4. In practice we get three price/quality indexes going from 0 to 1 so that the trade share assigned to the high-end segment increases with UVs, that assigned to the low-end segment decreases with UVs and that assigned to the middle-end segment increases up to r = 1 and decreases thereafter.
As a robustness check we used a simpler methodology, assigning each single trade flow as a whole to the respective segment according to its relative UV. This methodology is such that if r < 1/(1 + α) the product belongs to the low-end segment, if r > (1 + α) it belongs to the high-end segment, and if r is in between the two thresholds, it belongs to the middle-end segment. We used α = 0.25, meaning that in the middle-end segment UVs cannot differ more than 25% from the world average. This methodology is employed also in Fontagné-Freudenberg (1997) Intra-Industry Trade: Methodological Issues Reconsidered, CEPII.
The sophistication indexes for each price segment are computed as described in the paper on the trade flows that have been assigned to the respective segment.
1.3 A.3 Tables
Rights and permissions
Copyright information
© 2012 Springer-Verlag Berlin Heidelberg
About this chapter
Cite this chapter
Marvasi, E. (2012). The Sophistication of China’s Exports, Imports and Intermediate Products. In: Gomel, G., Marconi, D., Musu, I., Quintieri, B. (eds) The Chinese Economy. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-28638-4_9
Download citation
DOI: https://doi.org/10.1007/978-3-642-28638-4_9
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-28637-7
Online ISBN: 978-3-642-28638-4
eBook Packages: Business and EconomicsEconomics and Finance (R0)