View
4
Download
0
Category
Preview:
Citation preview
How Do Exporters Respond to AntidumpingInvestigations?
Yi Lua, Zhigang Taob and Yan Zhangb
aNational University of Singapore, bUniversity of Hong Kong
March 2013
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 1 / 45
IntroductionBackground
Increasing trade �ows due to rounds of reduction in tari¤s andadvancement in telecommunications and logistics
Yet persistent and even increasing use of contingent trade protectionpolicies (especially antidumping investigations), which are permissibleunder the World Trade Organization (WTO) rules and regulations(e.g., Prusa, 2001; Zanardi, 2006; Bown, 2011).
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 2 / 45
IntroductionResearch Question
Existing research generally focuses on the impact of antidumpingmeasures on protected �rms and industries (see, for example,Gallaway, Blonigen and Flynn (1999), Konings and Vandenbussche(2008), and Pierce (2011))
Limited studies on the impact of antidumping measures on a¤ectedforeign exporters.Using China Customs data that covers monthly transactions of allChinese exporters, we investigate how Chinese exporters respond toU.S. antidumping investigations during the period of 2000-2006.
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 3 / 45
IntroductionImportance
Essential for piecing up a picture of market competition betweendomestic �rms and foreign exporters in both the short-run and thelong-run, and its implications for industry dynamics and nationalcompetitiveness
Understanding whether foreign a¤ected exporters should continuetheir exporting behavior in response to negative shocks brought byantidumping investigations compliments the existing �rmheterogeneity literature that focuses primarily on the entry decisioninto export market
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 4 / 45
IntroductionOur Focus
Anatomize how the trade-dampening e¤ect operates: extensivemargin versus intensive margin
Investigate how di¤erent exporters (direct exporters versus tradeintermediaries, and single-product direct exporters versusmulti-products direct exporters) may respond to antidumpinginvestigations
Study whether exporters adjust export prices of the concernedproducts.
Examine di¤erential impacts across di¤erent stages of theantidumping investigation process
Finally, provide a coherent explanation to the aforementioned �ndingsbased on recent developments in trade theories.
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 5 / 45
Road Map
Background
Data
Estimation Strategy: DID
Main Results: Figures
Robustness
Discussion
Conclusion
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 6 / 45
Background
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 7 / 45
Data
China Customs Data over the period of 2000-2006
covers monthly export transaction of every Chinese exporter to theU.S., including product information (classi�ed at the Chinese HS-8 digitlevel), export volume, export value, and identity of Chinese exporters
Global Antidumping Database from the World Bank
has detailed information on each antidumping case, such as productinformation (classi�ed at the U.S. HS-10 digit level), initiation date,preliminary ITC and DoC determination dates, and �nal ITC and DoCdetermination dates
We match the two data sets at the HS-6 digit level, the mostdisaggregated level at which the two data are comparable
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 8 / 45
DataUS Antidumping Cases against Chinese Exporters over 2000-2006
A total of 47 U.S. antidumping cases against Chinese exporters
Two cases (one in early 2000 and the other in late 2006) are droppedas there is not enough pre- or post-antidumping period for us to carryout di¤erence-in-di¤erences estimation.
28 cases out of the remaining 42 cases ended up with a¢ rmative �nalITC determination (referred to as successful cases)
5 out of the 6 cases that had a¢ rmative preliminary ITCdetermination received negative �nal ITC determination (referred toas unsuccessful cases) and 1 was withdrawn before the �nal ITCdetermination (referred to as withdrawn cases);
8 cases were either withdrawn before the preliminary ITCdetermination or given the negative preliminary ITC determination(referred to as terminated cases).
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 9 / 45
Estimation strategy
To identify the possible e¤ects of antidumping investigations, weemploy the di¤erence-in-di¤erences (DID) estimation strategy at boththe product level (de�ned at HS-6 digit) and the �rm-product level
time variation: before and after the relevant stages of the antidumpinginvestigation process: initiation, ITC preliminary determination, andITC �nal determinationcross-sectional variation: a¤ected products (treatment group) anduna¤ected (control group)
Control groups
all una¤ected products/�rms within the same HS-4 digit productcategory where the a¤ected products/�rmsa matched group, constructed using the method of Blonigen and Park(2004).
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 10 / 45
Estimation strategy
Estimation speci�cation
binary variables
ypt = β1Treatmentp � Post1pt + β2Treatmentp � Post2pt+β3Treatmentp � Post3pt + λp + λt + εpt , (1)
duties
ypt = β1Treatmentp � Post1pt + β2Preliminary Dutiespt � Post2pt+β3Final Dutiespt � Post3pt + λp + λt + εpt , (2)
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 11 / 45
Main ResultsProduct-level Quantity Response
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 12 / 45
Main ResultsProduct-level Quantity Response
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 13 / 45
Main ResultsExtensive versus Intensive Margins: Extensive Margin
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 14 / 45
Main ResultsExtensive versus Intensive Margins: Extensive Margin
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 15 / 45
Main ResultsExtensive versus Intensive Margins: Intensive Margin
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 16 / 45
Main ResultsExtensive versus Intensive Margins: Intensive Margin
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 17 / 45
Main ResultsExtensive versus Intensive Margins: Intensive Margin
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 18 / 45
Main ResultsHeterogeneous Responses: Trade Intermediaries versus Direct Exporters
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 19 / 45
Main ResultsHeterogeneous Responses: Single-product versus Multi-product Direct Exporters
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 20 / 45
Main ResultsPrice Response
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 21 / 45
Main ResultsPrice Response
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 22 / 45
Main ResultsPrice Response
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 23 / 45
Main ResultsTrade-de�ection Response
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 24 / 45
Main ResultsTrade-de�ection Response
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 25 / 45
Robustness checks
Checks on the DID identi�cation assumption
di¤erential time trends before the antidumping investigation (Table A3)inclusion of product-speci�c time trends (Table A4)
Measurement errors
quarterly instead of monthly data (Table A5)exclusion of outliers � the top and bottom 1% observations (Table A6)
Inclusion of unsuccessful and withdrawn cases (Table A7)
Exclusion of antidumping cases concurrently investigated by othercountries (Table A8)
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 26 / 45
Robustness checks
Exclusion of processing trade (Table A9)
Exclusion of foreign �rms (Table A10)
Aggregation bias (Table A11)
Controlling for other trade shocks
safeguard measures (Table A12)China�s WTO accession (Table A13)
Di¤erent products (import demand elasticity) (Table A14)
Alternative de�nition of single-product direct exporters (Table A15)
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 27 / 45
DiscussionDi¤erences across Products
Summary of the �ndings
there is signi�cant extensive margin e¤ect, i.e., a sharp decrease in thenumber of exportersintensive margin e¤ect is not found when a binary variable of treatmentstatus is used, but uncovered when the antidumping duties are usedthere is little adjustment in F.O.B. export prices when a binary variableof treatment status is used, but a modest increase in prices when theantidumping duties are usedno trade de�ection
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 28 / 45
DiscussionDi¤erences across Products
No trade de�ection
to Canada or the EU (similar structure to the U.S.)among di¤erent types of exportersdi¤erent quantiles
Possible explanations
the �xed costs of exporting are country-speci�c (e.g., Chaney, 2008;Arkolakis, 2010)Indeed, we �nd in our data that Chinese exporters to the U.S. areheavily weighted in the U.S. market (63% of these exporters�worldexport revenues)
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 29 / 45
DiscussionDi¤erences across Products
Inference: products facing di¤erent levels of antidumping duties maybehave di¤erently
Further investigation: three quantiles
low (i.e., < 50%)medium (i.e., 50 � 100%)high (i.e., > 100%) antidumping duties.
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 30 / 45
DiscussionDi¤erences across Products
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 31 / 45
DiscussionDi¤erences across Products: Low Quantile
Low margins of antidumping duties are quite small: average about20.85%
the a¢ rmative determinations should be viewed as surprises to theproducers in these products
according to the study by Blonigen and Park (2004), exporters undersuch a scenario will raise prices over time
the small magnitude of duties makes it easy for producers easy raiseprices, and by doing so, they can get rid of the duties (and nuisances)in the future years through administrative reviewsraising prices is achievable given that these products are relativelyinelastic (i.e., the average elasticity of import substitution is 3.1)
The estimation results suggest that producers in these productsincrease prices by around 24%
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 32 / 45
DiscussionDi¤erences across Products: Low Quantile
Final sales prices of the export products concerned in the U.S. marketwill generally increase
lead to a decline in demand of the products concernedat the status quo, shrinking market demand is likely to lead to adecrease in �rm export volume across the boardindeed, it is found that the average export volume per exporterdecreases by 20%
Exporters could compensate the loss in demand by the increase intheir F.O.B. export prices
24% increase in export prices versus 20% decrease in export volumeno signi�cant exit of exporters
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 33 / 45
DiscussionDi¤erences across Products: High Quantile
High margins of antidumping duties: average about 185%
Such huge negative shocks drive weak exporters out of the market
especially, with large variations in exporter heterogeneity: i.e., theaverage coe¢ cient of variation is 0.32indeed, we �nd that the number of exporters fall by 52%
The wipeout substantially consolidate the market
more productive exporters can not just survive but even grow after theimposition of antidumping dutiesindeed, we �nd that surviving exporters modestly increase their F.O.B.prices and maintain their export volume by grabbing the market left theexiting �rms
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 34 / 45
DiscussionDi¤erences across Products: Medium Quantile
Blow is relatively large but not devastating: average about 86%
a number of exporters exit the market: fall by 30%
The post market remains competitive
especially in such elastic markets: the average elasticity of importsubstitution is 8.98surviving exporters may lower the prices to increase the competitivenessand their export volume also fall
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 35 / 45
DiscussionDi¤erences across Firms
Summary of the �ndings
less productive �rms more likely to exit U.S. marketdirect exporters more likely to exit U.S. market than are tradeintermediariesmulti-product direct exporters are more likely to exit the U.S. marketthan are single-product direct exporters upon issuance of an a¢ rmativepreliminary ITC determination, but the opposite holds following ana¢ rmative �nal ITC determination
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 36 / 45
DiscussionDi¤erences across Firms: Productivity E¤ect
Generally in line with the �rm heterogeneity literature
in the case of a per-period �xed cost of exporting (Melitz, 2003), thenegative shock causes a fall in export revenue, as a result of whichsome less productive ones are unable to recover the per-period �xedcost of exportingin the world without �xed cost of exporting (Melitz and Ottaviano,2008), the negative shock causes a decrease in exporters�markups, as aresult of which some less productive ones incur losses
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 37 / 45
DiscussionDi¤erences across Firms: Trade Intermediaries versus Direct Exporters
Stylized facts
trade intermediaries are more multi-market for the concerned products,and sell more products in the US68% of trade intermediaries sell the a¤ected products to countriesother than the U.S., versus 64% for direct exporters91% of trade intermediaries sell products other than the a¤ectedproducts in the U.S. market, versus 81% for direct exporterstrade intermediaries could tap into their reserves in other products andother markets to cross-subsidize their a¤ected products in the U.S.,which allows them to better weather the storm brought by theantidumping investigations
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 38 / 45
DiscussionDi¤erences across Firms: Single- versus Multi-product Direct Exporters
Overall, single-product direct exporters are more likely to exit the U.S.market
This can be explained by the greater capabilities of multi-productdirect exporters to cross-subsidize the a¤ected products than theirsingle-product counterparts, which is in line with our aforementionedexplanation on the di¤erential likelihood of exiting between tradeintermediaries and direct exporters
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 39 / 45
DiscussionDi¤erences across Firms: Single- versus Multi-product Direct Exporters
Di¤erence between preliminary and �nal determination: uncertainty
A simple model
two periods: 1 for preliminary and 2 for �nal determinationpro�ts in each period: πduties paid: dprobability of a¢ rmative �nal determination at period 1: p (28 out of34 in our sample)
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 40 / 45
DiscussionDi¤erences across Firms: Single- versus Multi-product Direct Exporters
Choices
period 2: if π > d , stay; if π < d , exitperiod 1: if stay, expected payo¤ isπ � pd + δ(π � pd) = (1+ δ)(π � pd); if exit, 0) if π > pd , stay; if π < pd , exit
Equilibrium
if π > d , stay after the preliminary and �nal determinationsif d > π > pd , stay after the preliminary but exit after the �naldeterminationif pd > π, exit after the preliminary
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 41 / 45
DiscussionDi¤erences across Firms: Single- versus Multi-product Direct Exporters
Our �ndings imply
single-product direct exporters relatively concentrate in the secondgroupmulti-product direct exporters relatively separate in the �rst and thirdgroupsmulti-product direct exporters are more heterogeneous thansingle-product direct exporters
Dispersion in the data
single-product direct exporter: 0.23multi-product direct exporter: 0.32
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 42 / 45
Conclusion and Implications
Our results suggest that U.S. antidumping investigations wipe outweaker Chinese exporters and leave behind more productive exportersoften with multi-market and multi-product coverage
In many product categories (especially those facing high margins ofantidumping duties), the wipeout results in a substantialconsolidation of Chinese exporters, under which the surviving,stronger Chinese exporters can even raise their F.O.B. export pricesand at the same time grab the market share left by the weaker ones
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 43 / 45
Conclusion and Implications
Existing studies (e.g., Pierce, 2011) on the impacts of U.S.antidumping measures on its domestic, protected �rms have shownthat while protected �rms are able to increase their prices, theirphysical productivity actually falls
And the protection through temporary imposition of antidumpingduties is more tilted toward the weaker domestic producers, therebyslowing down the resource reallocation towards more productive ones.
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 44 / 45
Conclusion and Implications
Taken together, U.S. antidumping investigations de�nitely bringtemporary bene�ts to domestic producers, who expand their marketshare, as Chinese imports substantially fall and numerous Chineseexporters exit the market
In the long-run (especially when the antidumping duties are lifted),however, antidumping investigations may spell more troubles for U.S.domestic producers in their competition with the Chinese exporters,as the former become less productive on average whereas the latterexperience just the opposite
Lu, Tao, Zhang (NUS, HKU) How Do Exporters Respond to Antidumping Investigations? March 2013 45 / 45
Recommended