CAHIER DE RECHERCHE #1621E Département de science économique Faculté des sciences sociales Université d’Ottawa
WORKING PAPER #1621E Department of Economics Faculty of Social Sciences University of Ottawa
The Persistence of Trade Policy in China After WTO Accession*
November 2016
Copyright By PowCoder代写 加微信 powcoder
*I thank , and for their valuable input and support, along with Bøler, Francisco Costa, Réka Juhász, , , , , and Qinghua Tang, and discussants and seminar participants at the London School of Economics, Bank of Canada, McGill, Ottawa, Waterloo, Alberta School of Business, Oxford, the Northeast Universities Development Consortium (NEUDC) Conference, the China Economics Summer Institute, the Annual Conference of the Canadian Economics Association, the Centre for Economic Performance Annual Conference and the Globalisation and Economic Policy Postgraduate Conference.
† Department of Economics, University of Ottawa, 120 University Private, Ottawa, Ontario, Canada, K1N 6N5; E- mail:
Import tariffs have fallen steeply worldwide over the last several decades, but has trade policy persisted through a rise in the use of other instruments? I study this question in the context of China’s 2001 accession to the World Trade Organization, using panel data on Chinese export policies. I find that after its entry into WTO, the distribution of China’s export restrictions across industries increasingly resembles the inverse of its pre-WTO import tariff schedule. The evidence suggests that increases in export restrictions are likely to have partly restored China’s pre-WTO pattern of industrial protection.
JEL Classification: F13, F14, O13, O24.
1 Introduction
A striking stylized fact about the international economy of the last several decades is the dramatic worldwide decline in the most widely observed instrument of trade policy: the import tariff. Much of this decline has been credited to the GATT/WTO process of multi- lateral trade negotiations, in which governments have committed to the implementation of ever smaller tariffs. But does the demise of the import tariff signal a retreat from activist trade policy among governments, or has trade policy simply persisted in other forms instead?
Long-established results in international trade theory suggest that other policy instru- ments may partially or entirely reproduce the effects of import tariffs. This may be ac- complished even with policies that only indirectly affect imports; for instance, symmetries between the effects of export taxes and import tariffs in general equilibrium have been known to economists at least since the work of Lerner (1936). Yet while we have excellent data on import tariffs across products, countries and time, there is little systematic information on countries’ usage of the other instruments that could substitute for tariffs in the implemen- tation of their trade and industrial policies. This gap in the data leaves us unable to fully judge the extent to which such policies remain an active feature of the global economy, or establish the effectiveness of GATT/WTO negotiations in actually changing governments’ behaviour.
This issue is of particular relevance to developing countries, for whom the instruments of trade policy, especially import tariffs, have historically been important tools of industrial strategy. While some observers have suggested that there is still much scope for less de- veloped countries to implement industrial policies in the current multilateral system (e.g. Rodrik 2004), others have portrayed multilateral trade agreements as overly restrictive in this respect (e.g. Chang 2002). Here, I consider whether participation in WTO, and the accompanying restrictions on policy choice, have served as an effective constraint on the pattern of industrial protection of China – a developing country with a well-known history of interventionist economic policy, and the most important recent entrant into WTO.
China’s 2001 entry into WTO allowed it to benefit from improved access to foreign markets, including permanent most-favoured-nation status in the United States. At the same time, in order to secure the assent of incumbent members to its WTO accession, China was also required to agree to changes in its own policies. For example, like other WTO members, China pledged as a condition of its accession to permanently keep its import tariffs below agreed maximum levels. This required cuts to its pre-WTO tariffs, and a resulting change in the pattern of protection across Chinese industries.
However, although these tariff cuts have been successfully completed, China has recently been the subject of trade disputes involving several other policies, such as domestic content requirements, preferential loans and discriminatory tax treatment. This raises the question of whether such policies may have partly restored China’s pre-WTO pattern of industrial protection. As noted above, economic theory suggests that China could use various instru- ments to achieve trade or industrial policy goals previously accomplished with import tariffs. Here, I consider one particularly important class of instruments for which comprehensive and readily quantifiable data on Chinese policies is available: export restrictions.
Figure 1: Standard deviation of China’s applied and bound import tariffs and export tax equivalents of export VAT rebate policies and export duties across nonagricultural products, 1997 to 2012
In particular, I gather panel data on several instruments for which, like import tariffs, product-level policy schedules are published frequently in China. Figure 1 plots the standard deviation across products of import tariffs and the export tax equivalents of two of these instruments over time. As shown in the figure, I find that compression in China’s tariff schedule due to its WTO accession commitments has been followed by a rise in variation in export restrictions across products. Moreover, I document in the empirical analysis below that these post-accession changes in China’s export restrictions are systematically related to its pre-WTO schedule of import tariffs.
I begin by observing that industries protected by higher import tariffs before WTO accession were subsequently given relatively greater export support via smaller increases in export restrictions. Similarly, post-accession rises in export restrictions are also negatively related to China’s pre-WTO non-tariff import barriers. This suggests that in the presence of intra-industry trade, China’s export policies support a similar set of industries to those favoured by its pre-WTO pattern of import protection.
I then consider inter-industry relationships, showing that while the sectors with the largest pre-WTO tariffs were producers of goods relatively downstream in the value chain, China’s subsequent export tax rises have been concentrated in raw materials industries. In theory, by imposing export taxes on raw materials, an exporter of these goods (such as China, which is a major producer of a wide range of raw materials) can generate a wedge be- tween their domestic and world prices, supporting domestic downstream firms by providing them with an input cost advantage. Indeed, the US and EU have twice taken China to the
WTO’s dispute settlement mechanism over its export restrictions on raw materials, with the claim that these are used to support downstream industries. I find that downstream sectors indirectly protected through relatively lower input tariffs before WTO accession were subse- quently supported via larger increases in export taxes on inputs from the same industries.
Finally, I examine whether the recent changes in China’s export restrictions have been associated with changes in Chinese export patterns. As expected, I observe a negative relationship between export taxes and exports, along with a positive relationship between upstream export taxes and exports of downstream products. I also find that industries subject to larger export restrictions have a lower ratio of exports to total sales, suggesting that sales of local raw materials have been diverted to the domestic market.
These findings may be interpreted in two closely related ways. First, by requiring China to bind its import tariffs below certain levels, WTO accession restricted China’s flexibility to use across-industry variation in import policy to protect some industries relative to others. However, trade policy has nonetheless persisted in China through increasing variation across industries in export restrictions, which have tended to support the same industries that benefited from tariff protection before China’s WTO accession. Second, since the tariff reductions required as conditions of China’s accession are highly correlated to its pre-WTO tariff levels, China’s recent export policy changes have partly offset these tariff cuts. Either interpretation suggests that a full assessment of the changes in China’s trade policy after its entry into WTO must take instruments other than import restrictions into account.
While this study is concerned with a single country and event, the potential for substitu- tion between instruments of trade policy is relevant well beyond China’s WTO accession. As early as 1984, Baldwin suggested in the first Handbook of International Economics that non- tariff barriers “have been used more extensively by governments to attain the protectionist goals formerly achieved with tariffs.” Since then, a handful of other empirical studies have found evidence of substitution from import tariffs to other policies in developing countries including India (Bown and Tovar 2011) and Turkey (Lima ̃o and Tovar 2011), as well as in the US (Ray and Marvel 1984) and across countries (Bown and Crowley 2014).1 However, these studies all consider substitution between import tariffs and other import-side measures (such as anti-dumping duties); the study of substitution between import-side and export-side policies in the same country is an innovation of this study.2
In fact, although there is a long theoretical literature on symmetries between import tariffs and export taxes, beginning with Lerner’s classic contribution in 1936,3 very few papers have actually observed an empirical relationship between tariffs and export taxes.4
1See Feinberg and Reynolds (2007), Vandenbussche and Zanardi (2010) and Moore and Zanardi (2011) for other cross-country analyses. Anderson and Schmitt (2003) present a theoretical study of substitution between import-side policies.
2Some observers have drawn parallels between cuts in import tariffs in one country and the negotiation of agreements requiring its trading partners to place quotas on exports (‘voluntary export restraints’, or VERs). See Yu (2000) for a theoretical discussion.
3For example, McKinnon (1966) provides an theoretical extension of Lerner symmetry to an economy with intermediate goods.
4Golub and Finger (1979) observe a cross-country relationship of this kind, noting parallels between import tariffs on downstream goods in developed countries and export taxes on raw materials in their less
This dearth of empirical papers on a long-studied theoretical topic is partly due to the fact that global data on export taxes is sparse. The assembly of detailed panel data on export restrictions in China, allowing for comparisons with trends in import tariffs, is thus a contribution to this literature.5,6
The relationship between the trade policies studied here and the value chain also situates this paper in a literature linking growth in exports of manufactures to advantageous access to raw materials. In particular, parallels with the case of the United States around the turn of the twentieth century, as studied by Wright (1990) and Irwin (2003), are notable. Wright finds that intensive exploitation of a wide range of local resource endowments played an important role in US manufacturing success during this period, while Irwin reaches similar conclusions in a more focused study of the American iron and steel industry. According to Irwin, the rise in US exports of iron and steel around 1900 was driven in part by the exploitation of a large new deposit of iron ore, whose output was kept within the US due to high transportation costs and vertical integration. Like the US at that time, present-day China is a global leader in the production of many raw materials, but my analysis suggests that privileged access to those inputs for Chinese manufacturers has been generated, at least in part, by state-imposed restrictions on their export.
Finally, because of China’s size and the importance of its WTO accession to the world economy, studies of the effects of China’s WTO entry such as this one are of particular interest to scholars of both economic development and international trade. Existing papers have focused mainly on the effects of tariff cuts on local outcomes (e.g. Chen and Ravallion 2004, Brandt et al. 2012a), rather than their implications for local policies. The apparent presence of policy substitution in China suggests that these and other studies might underestimate the direct effects of China’s tariff cuts if they do not take domestic policy responses into account.
The remainder of the paper is laid out as follows. Section 2 provides background informa- tion on China’s WTO accession, including its commitments regarding import tariffs. Section 3 introduces the export restrictions data and discusses the evolution of Chinese export re- strictions over time. Section 4 analyzes the relationship between China’s export restrictions and its import tariffs, while Section 5 examines whether export restrictions have actually affected China’s export patterns. Section 6 then draws conclusions.
developed trading partners. See also Latina et al. (2011).
5This data is available to other researchers on my web site.
6In work contemporaneous to this study, Eisenbarth (2014) and Gourdon et al. (2016a) use data on some
of the export restrictions studied here to investigate the motivation for these policies, but neither of these papers explores their implications for the effects of import tariffs. The dataset in this paper is also distinctive in that it includes a wider range of policy instruments and years. Chandra and Long (2013) calculate the elasticity of Chinese exports to VAT rebate rates using firm-level data on VAT payments, and Gourdon et al. (2016b) also estimate the effects on exports of China’s VAT rebate policy, though neither of these papers examines the indirect effects of upstream policies on downstream exports. Solleder (2013) compiles data on export taxes from twenty countries, including two years of data on export duties in China.
2 Background: China and WTO
After finalizing WTO accession agreements with the United States in 1999 and the European Union in 2000, China entered WTO in December 2001. China already held most-favoured- nation (MFN) status in each of its main trading partners at the time of its entry into WTO, so the schedule of import tariffs that it faced in these countries did not change after its WTO membership. However, China gained market access abroad through a decline in trade policy uncertainty: perhaps most importantly, China’s MFN status in the US was subject to annual renewal before 2001, but was made permanent upon China’s accession to WTO.7
As a WTO member, China became bound both by WTO rules and by additional specific commitments made as conditions of its accession. Like other countries joining WTO, one of China’s key commitments was to permanently set its import tariffs at or below levels agreed in international negotiations. For almost all products, this bound tariff rate was equal to or smaller than China’s applied tariff in 1999, the year in which agreement on tariffs on industrial products was reached, and so tariff cuts were required in order to meet this condition.8 The schedule for implementation of China’s bound tariffs extended to 2010, with most tariffs to be reduced to their bound rates by 2005.
In practice, China’s nonagricultural applied tariffs were indeed reduced to their bound levels after its WTO accession. As a consequence, China’s mean applied tariff across nona- gricultural products decreased from 16% in 1999 to 9% in 2012, as shown in Table 1 Panel A. Because negotiated tariff cuts were highly correlated to initial tariff levels (with a cor- relation coefficient of approximately 0.8 across nonagricultural products), these cuts led to a compression of China’s tariff schedule. Panel A of Table 1 accordingly shows that the standard deviation of China’s applied tariffs across products declined along with the mean.
Since its accession to WTO, a number of China’s policies have been the subject of disputes addressed through the WTO’s dispute settlement mechanism. During the period studied in this paper (up to 2012), China was the respondent in 19 separate cases brought by other WTO members. Two of these disputes involved its restrictions on exports: China – Measures related to the exportation of various raw materials (brought to the dispute settlement mech- anism by the US, EU and Mexico in 2009) and China – Measures related to the exportation of rare earths, tungsten and molybdenum (brought by the US, EU and Japan in 2012).
The complainants in these two cases portrayed China’s export restrictions as important instruments of industrial policy; a US government submission for the 2009 case argued that “China’s industrial strategy is to leverage and exploit the differences in the international and
7Handley and Lim ̃ao (2013) and Pierce and Schott (2016) find that uncertainty related to the differ- ence between US MFN tariffs and the non-MFN tariffs that would otherwise have prevailed for China has explanatory power for the evolution of US manufacturing employment after 2001.
8Throughout the paper, I restrict the analysis to nonagricultural products (or industries), and also omit important agricultural inputs (fertilizers and pesticides); see the data appendix for details. This is because liberalization of trade in agricultural products, where nontariff barriers tend to be particularly important, often involved replacement of nontariff barriers with tariff-based protection during this period rather than tariff cuts (Branstetter and Lardy 2008). The inclusion of agricultural products into the analysis therefore complicates the interpretation of the relationship between export policies and import tariffs in Section 4. Including agricultural products somewhat weakens but does not otherwise change the results in the paper.
Table 1: Summary statistics of Chinese trade policies
Panel A. Import tariffs
Import tariffs .161 (.098)
Bound (.060)
Panel B. Export VAT rebate policies and export duties
Export tax equivalent of export VAT rebate policies
Export duties
Joint export tax equivalent
2002 2007 2012 .019 .057 .063
Panel C. Export processing prohibitions
2005 Export processing prohibitions .0003
Panel D. Other export restrictions
2007 2012 .062 .158
Export licenses
Export quotas
State or designated trading Any of the above
2002 2007 2012 .020 .039 .042 .019 .014 .017 .008 .007 .007 .022 .041 .043
1999 (applied)
2012 (applied)
(.020) (.040)
.002 .004 .007
(.023) (.026)
.021 .063 .073
(.033) (.063)
(.036) (.093)
These summary statistics include the mean and standard deviation (in brackets) across nonagricultural products of the policies in Panels A and B and the proportion of nonagricultural products covered by the policies in Panels C and D. See the data appendix for information on data sources for all
程序代写 CS代考 加微信: powcoder QQ: 1823890830 Email: powcoder@163.com