Does Yuan Appreciation Weaken the Increase in Exporters due to Trade Liberalization? Evidence from Chinese Firm-Product Data Zhe Chen∗
Yoshinori Kurokawa†
Preliminary Draft: September 8, 2017
Abstract Using Chinese firm-product data from 2000 to 2006, this paper empirically tests whether or not the appreciation (depreciation) of Chinese Yuan weakens (strengthens) the effect of trade liberalization on the extensive margin of China’s exports to 170 countries. Based on regressions, we have four main empirical findings. First, reductions in tariffs, charged by China’s trade partners, increased China’s exporter numbers and export value/quantity per exporter at the product level while the appreciation (depreciation) of Chinese Yuan decreased (increased) them—the effect of exchange rates is larger than that of tariffs in all cases. Second, reductions in tariffs, charged by trade partners, do not affect the entry or exit of China’s exporters while Chinese Yuan appreciation/depreciation does affect entry but does not affect exit. Third, processing trade firms are more affected by tariffs and exchange rates than ordinary trade firms are. Fourth, the effects are more significant if export destinations are non-OECD countries.
JEL classification: F12, F14, F31 Keywords: Melitz model, exchange rates, tariffs, extensive margin, exports
∗
School of International Trade and Economics, University of International Business and Economics. 1302 Boxue Building, 10 Huixin Dongjie, Chaoyang District, Beijing 100029, China. E-mail:
[email protected]. † Faculty of Humanities and Social Sciences, University of Tsukuba. 1-1-1 Tennodai, Tsukuba, Ibaraki 305-8571, Japan. E-mail:
[email protected].
1
1
Introduction
One of the most notable works in the recent history of international trade is the Melitz (2003) model of heterogeneous firm trade. The model theoretically shows that a reduction in trade costs decreases the export cutoff productivity level, increasing the number of exporters—the extensive margin (henceforth EM) of exports. In fact, it has been empirically documented that the EM indeed increased after trade liberalization (e.g., Kehoe and Ruhl, 2013). Little attention, however, has been paid to the effects of nominal exchange rates in this literature.1 This is probably because the Melitz model of trade is a real model and thus does not explicitly deal with the monetary aspects of trade such as changes in nominal exchange rates, although it is possible to extend the discussion of the Melitz model to analyze the effect of nominal exchange rates. In fact, introducing nominal exchange rates (the price of the foreign currency in the home currency) into the Melitz model implies the following:2 Under the fixed exchange rate regime (say, nominal exchange rates are constant), the model is the same as the original the Melitz model. Thus, as is well-known, a reduction in trade costs decreases the export cutoff productivity level, increasing the EM of exports. Under the flexible exchange rate regime, however, if a reduction in trade costs is weakened by a change in nominal exchange rates (appreciation), the export cutoff moves less and thus the EM increases less than if only trade costs reduce. On the other hand, if the effect of a reduction in trade costs is strengthened by a change in nominal exchange rates (depreciation), the export cutoff moves more and thus the EM increases more than if only trade costs reduce.3 Thus, while it has been empirically documented that the EM indeed increased after trade liberalization, the observed changes in EM after trade liberalization might have been the results weakened or strengthened by changes in nominal exchange rates. In other words, they might have been stronger or weaker without changes in nominal exchange rates. 1
As will be mentioned later, Bergin and Lin (2008) examine the effect of exchange rate regimes on the EM of trade. 2 Deriving the implications on the EM from the Melitz model with nominal exchange rates is simple and has been relegated to Appendix A.1. Note that although they derive a different empirical question from ours, Baggs et al. (2009) also introduce exchange rates into the Melitz model. Their question is whether the effects of real exchange rates on firm survival/exit and sales are different for firms with different productivity. 3 Feenstra (1989) theoretically argues that the effects of tariffs and exchange rates on the price of imports are symmetric, and empirically tests whether their effects on U.S. prices of Japanese cars, trucks, and motorcycles are symmetric.
2
This is important particularly for empirical studies on trade liberalization and the EM because they might have underestimated or overestimated the pure effect of trade liberalization on the EM. To the best of our knowledge, however, no previous studies have empirically tested the above implications from the Melitz model.4 We now fill this void. We use Chinese firm-product data from 2000 to 2006 to test whether or not the appreciation (depreciation) of Chinese Yuan weakens (strengthens) the effect of trade liberalization on the EM of China’s exports to 170 countries. The benchmark regression results suggest that reductions in tariffs that are charged by China’s trade partners increased China’s exporter numbers at the product level while the appreciation (depreciation) of Chinese Yuan decreased (increased) them. The results thus support the theoretical implication that an increase in the EM of exports due to trade liberalization can be weakened by currency appreciation or strengthened by currency depreciation. The regression results are similar for the export value and quantity per exporter—the intensive margin of exports—at the product level. In all cases, the impact of nominal exchange rates is larger than that of tariffs. When we investigate the effects on the entry and exit of China’s exporters, the regression results suggest that reductions in tariffs, charged by trade partners, do not affect the entry or exit of exporters while Chinese Yuan appreciation/depreciation does affect entry but does not affect exit. We also perform several robustness checks regarding trade modes (ordinary vs. processing trade), export destinations (OECD vs. non-OECD countries), and real exchange rates. Thus, we make the following contributions to the literature on the EM of trade. One set of studies (e.g., Melitz, 2003) focus on the effect of trade liberalization on the EM of trade. Another set of studies (e.g., Bergin and Lin, 2008) focus on the effect of exchange rates on the EM of trade. In this line of studies, we focus on the effects of these two factors on the EM of trade, in particular, using Chinese firm-product data we empirically investigate whether or not the effect of trade liberalization on the EM of China’s exports is weakened (strengthened) by Chinese Yuan appreciation (depreciation). Baggs et al. (2009) (henceforth BBF) also look at both liberalization and exchange rates, but they empirically test, using Canadian industry4
Note that as mentioned by Baggs et al. (2009), most empirical studies that examine tariff changes/trade liberalization include exchange rates as a control variable; however, the analysis is at the industry-level, not at the firm-level.
3
firm data, if tariffs and real (not nominal) exchange rates symmetrically affect Canadian firm survival/exit (not the EM of exports). Moreover, we have recently found Fitzgerald and Hallerz (2017) (henceforth FH), who empirically test, using Irish firm-product data, if tariffs and real exchange rates symmetrically affect Irish exporters. Their paper is close to BBF in that both empirically test the equivalence between tariffs and real exchange rates; however, it is also close to our paper in that both focus on the effects on exporters. FH are thus somewhere in the middle of BBF and our paper. There are four major differences among our paper, BBF, and FH. The first difference is that while BBF and FH both focus on real exchange rates, our paper focuses on nominal exchange rates, although we also do robustness checks with real ones. The second is that while BBF use industry-firm data, our paper and FH both use more detailed data, that is, firm-product data. The third is that while BBF and FH, respectively, cover the United States and 30 countries as trade partners, our paper covers much more countries, 170 countries. The fourth is that all of the three studies show different results: the effect of exchange rates is larger than that of tariffs in China (our paper), similar in Canada (BBF), and smaller in Ireland (FH). Therefore, these three papers are very good complements. The rest of this paper is organized as follows. Section 2 documents the regression specification and data. We report regression results in Section 3. Section 4 concludes.
2 2.1 2.1.1
Regression Specification and Data Regression Specification Exporter Numbers and Export Value/Quantity per Exporter
As has mentioned in the introduction, the Melitz model with nominal exchange rates implies that an increase in the EM of exports due to trade liberalization can be weakened by currency appreciation or strengthened by currency depreciation. Guided by this theoretical implication, we empirically investigate the impact of tariffs and exchange rates on exporter numbers at the product level. The benchmark regression is as follows: ln(Eijt ) = α0 + α1 ln(T arif fijt ) + α2 ln(N ERjt ) + Other Controls + ijt
4
(2.1)
Here, Eijt is the number of exporters of product i from China to country j at time t. T arif fijt is the tariff of product i charged by country j at time t. N ERjt is the nominal exchange rate, which is defined as Chinese Yuan per currency of country j at time t. Thus, when N ER increases, Chinese Yuan depreciates. Some inputs that are used by Chinese firms are imported, and thus we control China’s import tariffs within the same industry (HS2 level). Finally, we also control the product, country and time fixed effects. If α1 is negative and α2 is positive, then the implication from the Melitz model is supported. When T arif f decreases, the EM of exports increases. When Chinese Yuan appreciates (depreciates), the EM of exports decreases (increases). That is, the increase in EM of exports due to tariff reductions is weakened by currency appreciation or strengthened by currency depreciation. We also empirically investigate the impact on export value/quantity per exporter—the intensive margin of exports—at the product level. Then Eijt is the export value/quantity per exporter of product i from China to country j at time t. 2.1.2
Entry and Exit of Exporters
In the previous section, we have seen a change in exporter numbers which is the result of entry and exit. Thus we also investigate the impacts of tariffs and exchange rates on the entry and exit of exporters. The benchmark regression is as follows: ln(ENijt )d = β0 + β1 ln(T arif fijt )d + β2 ln(N ERjt )d + Other Controls + ηijt
(2.2)
Here, ln(ENijf t )d is the log difference of new exporter numbers or exit exporter numbers of product i to country j between year t − 1 and t. ln(T arif fijt )d is the log difference of tariff of product i charged by country j between year t − 1 and t. ln(N ERjt )d is the log difference of Chinese Yuan per currency of country j between year t − 1 and t. Hence, when ln(N ERjt )d increases, Chinese Yuan depreciates. We control the log difference of China’s import tariffs within the same industry (HS2 level). Finally, we also control the product, country and time fixed effects. When EN is new exporter numbers, if β1 is negative and β2 is positive, then a decrease in T arif f causes more Chinese firms to enter foreign markets while Chinese Yuan appreciation (depreciation) causes less (more) Chinese firms to enter. When EN is exit exporter numbers, if β1 is positive and β2 is negative, then a decrease in T arif f causes less
5
Chinese firms to quit foreign markets while Chinese Yuan appreciation (depreciation) causes more (less) Chinese firms to quit.
2.2
Data
The exchange rate and CPI data are from the IMF at the quarterly level.5 The tariff
6
data
is from the UNCTAD - Trade Analysis Information System (TRAINS). We have the import tariffs of each HS 6-digit level product for each country at the annual level. The Chinese export data is from Chinese Customs at the quarterly level. We have the number of exporters and the export value/quantity of each HS6 product. The sample period is from 2000/01 to 2006/12. The data summary is shown in Table 1 and Figure 1. The first part of Table 1 shows that the exporter and destination numbers increased over time. From 2000 to 2006, exporter numbers increased by 174.5% and destination numbers increased by 43%. Product numbers are almost constant. The second part of Table 1 shows that the tariffs, charged by partners, decreased over time. In 2000 the simple average tariffs are 9.6%, and in 2006 the tariffs are 8.1%. In addition, the simple average tariffs of OECD countries remained robust, and that of non-OECD countries decreased almost 30% from 2000 to 2006. The last part of Table 1 shows that China’s import tariffs also decreased from 2000 to 2006. Especially, when China became a WTO member in 2002, China’s average import tariffs decreased about 27% in that year. Figure 1 shows the time trend of China’s nominal effective exchange rates (NEERs) from 2000 to 2016. After China entered WTO in 2002, the NEERs decreased until 2005— depreciation. On July 21, 2005, the People’s Bank of China announced a revaluation of Chinese Yuan and a reform of the exchange rate regime. After 2005, the NEERs began to increase—appreciation—and increased until 2015. Note that here, an increase (decrease) in NEERs means China’s appreciation (depreciation) while an increase (decrease) in the independent variable N ER—defined as Chinese Yuan per a foreign currency—means Yuan depreciation (appreciation). 5 6
We also show results with monthly data in Appendix A.3. The tariff is the effectively applied tariff (AHS) which is defined as the lowest available tariff.
6
2.3
Tariffs and Exchange Rates
In this section, we present more detailed data for tariffs and nominal exchange rates, in particular, the time trends of tariffs and nominal exchange rates for China’s top 10 trade partners from 2000 to 2006, as shown in Figures 2 - 3. The top 10 partners are the euro area, the United States, Hong Kong, Japan, South Korea, the United Kingdom, Singapore, Taiwan, Canada, and India. Here, the tariff data is at the annual level, and the exchange rate data is at the monthly level. The figures imply that both tariffs and nominal exchange rates might be important factors contributing to changes in the EM of China’s exports. They also imply that the increase in EM due to tariff reductions might be weakened by currency appreciation or strengthened by currency depreciation. In the case of the euro area, for example, tariffs, charged by the euro area, decreased and Yuan depreciated from 2002 to 2004, implying that currency depreciation might have strengthened the increase in EM due to tariff reductions. In the case of Japan, tariffs, charged by Japan, decreased and Yuan first appreciated and next depreciated during the period 2001 to 2005, implying that currency appreciation and depreciation, respectively, might have weakened and strengthened the increase in EM due to tariff reductions.
3 3.1
Regression Results Exporter Numbers and Export Value/Quantity per Exporter
The column 1 of Table 2 shows the extensive margin effects of tariffs and nominal exchange rates. According to the Melitz model, a reduction in tariffs that are charged by foreign countries would encourage more Chinese firms entering foreign markets while the appreciation (depreciation) of Chinese Yuan would discourage (encourage). Our results verify this hypothesis. If the tariffs decrease by one percent, exporter numbers would increase by 0.04 percent. If Chinese Yuan appreciates by one percent, exporter numbers would decrease by 0.07 percent. From 2000 to 2005, the tariffs decreased and Chinese Yuan depreciated, and thus both two effects could stimulate the EM of exports. But after 2005, Chinese Yuan began to appreciate, which could weaken the effect of tariffs reductions. Especially, the impact of nominal exchange rates is larger than that of tariffs.
7
In addition, the column 1 of Table 2 also shows that the reductions in tariffs, charged by foreign countries, stimulate the intensive margin of exports from China and the appreciation (depreciation) of Chinese Yuan would weaken (strengthen) this effect. If the tariffs decrease by one percent, the export value per exporter would increase by 0.06 percent. Although the impact of tariff reduction is significant, the size is small. On the other hand, the impact of exchange rate fluctuations on exports is larger. If Chinese Yuan appreciates by one percent, the export value per exporter would decrease by 0.10 percent. The effects are similar for export quantity per exporter. If the tariffs decrease by one percent, export quantity per exporter would increase by 0.07 percent. If Chinese Yuan appreciates by one percent, export quantity per exporter would decrease by 0.09 percent. In the column 2, we additionally control the import tariffs charged by China. The effects of tariff reductions and exchange rate fluctuations remain robust. We also find that the import tariff reductions stimulate exports from China, at both the intensive and extensive margins. The reductions of import tariffs could decrease the input costs of Chinese firms, which improves their competitiveness in foreign markets. In sum, from 2000 to 2005, the reductions of tariffs charged by partners, the Chinese Yuan depreciation, and China’s import tariff reductions could all stimulate the exports of China, at both the intensive and extensive margins. But after 2005, the Chinese Yuan appreciation could weaken the benefits from trade liberalization.
3.2
Entry and Exit of Exporters
In this section, we investigate exporter dynamics in more details. Exporter numbers depend on both entry and exit of exporters. The tariffs and exchange rates might have heterogeneous effects on exporters’ entry and exit decisions. Thus, we separately examine the impacts. The column 1 of Table 3 shows that the reductions in tariffs, charged by foreign countries, would not affect the entry or exit of exporters. The Chinese Yuan appreciation (depreciation), on the other hand, would prevent (encourage) new firms entering foreign markets but has no effect on the exit of exporters. If Chinese Yuan appreciates by one percent, new exporter numbers would decrease by 0.32 percent. In the column 2, we additionally control China’s import tariffs, and the above pattern remains robust.
8
3.3
Robustness Checks
In this section, we examine whether trade modes, export destinations and real exchange rates would affect the benchmark results. Trade Modes: Processing trade refers to the activity of importing all or part of the raw and auxiliary materials from abroad, and re-exporting the finished products after processing or assembly by firms within the mainland. Thus, processing trade firms are more likely to be affected by the tariff and nominal exchange rate fluctuations. We thus divide firms into two groups: ordinary and processing trade firms. The columns 3 and 4 of Table 2 show that the effects on exporter numbers and export value/quantity per exporter of tariffs and nominal exchange rates are indeed larger for processing trade firms. However, the columns 3 and 4 of Table 3 show that effects on entry and exit are not clear. As can be seen, the effect of nominal exchange rates on new exporter numbers is more significant for ordinary trade while the effect on exit exporter numbers is more significant for processing trade. Export Destinations: After China became a WTO member in 2002, the tariffs charged by foreign countries, especially by developing countries, decreased a lot. For example, it can be seen by comparing India with developed countries (see Figures 2 and 3). Thus, we divide the export destinations into two groups: OECD and non-OECD countries. The columns 5 and 6 in Tables 2 and 3 show that the effects are more significant for non-OECD countries. Real Exchange Rates: Nominal and real exchange rates show very different movements in the United States over 2000-2006 (see Figure A1 in Appendix A.2). Motivated by this observation, we perform robustness checks with real exchange rates. Tables 4 and 5 show the results with real exchange rates. As can be seen, the results remain robust.
4
Conclusion
We have empirically tested the implications from the Melitz model with nominal exchange rates. Specifically, using Chinese firm-product data from 2000 to 2006 with 170 trade partners, we have tested whether or not the effect of trade liberalization on the EM of China’s exports is weakened (strengthened) by Chinese Yuan appreciation (depreciation). Based on regressions, we have four main empirical findings. First, reductions in tariffs, charged by China’s trade partners, increased China’s exporter numbers and export value/quantity 9
per exporter at the product level while the appreciation (depreciation) of Chinese Yuan decreased (increased) them—the latter effect is larger than the former one in all cases. Second, reductions in tariffs, charged by trade partners, do not affect the entry or exit of China’s exporters while Chinese Yuan appreciation/depreciation does affect entry but does not affect exit. Third, processing trade firms are more affected by tariffs and exchange rates than ordinary trade firms are. Fourth, the effects are more significant if export destinations are non-OECD countries. The results presented in this paper are important particularly for empirical studies on trade liberalization and the EM of trade. Because the results indicate that the observed increases in EM after trade liberalization could be the results weakened by currency appreciation or strengthened by currency depreciation, past studies might have under- or overestimated the pure effect of trade liberalization on the EM. Thus future studies will need to more carefully control exchange rates when investigating the changes in EM using firm data. Finally, our empirical results pose a challenge to theorists developing models that can explain why the effect of exchange rates on Chinese exporters is larger than that of tariffs. We leave it to future research.
10
References Baggs, Jen, Eugene Beaulieu, and Loretta Fung (2009) ‘Firm survival, performance, and the exchange rate.’ Canadian Journal of Economics 42(2), 393–421 Bergin, Paul R., and Ching-Yi Lin (2008) ‘Exchange rate regimes and the extensive margin of trade.’ NBER Working Paper 14126 Feenstra, Robert C. (1989) ‘Symmetric pass-through of tariffs and exchange rates under imperfect competition: An empirical test.’ Journal of International Economics 27(1-2), 25– 45 Fitzgerald, Doireann, and Stefanie Hallerz (2017) ‘Exporters and shocks.’ Federal Reserve Bank of Minneapolis Staff Report 549 Kehoe, Timothy, and Kim Ruhl (2013) ‘How important is the new goods margin in international trade?’ Journal of Political Economy 121(2), 358–392 Melitz, Marc J. (2003) ‘The impact of trade on aggregate industry productivity and intraindustry reallocation.’ Econometrica 71(6), 1695–1725
11
Table 1: Data Summary Year Variables
2000
2001
2002
2003
2004
2005
2006
Exporter Numbers
59,709
65,505
74,882
90,705
114,768
138,881
163,923
Product Numbers
4,789
4,809
4,778
4,816
4,853
4,879
4,894
104
134
141
124
125
135
149
(mean)
9.60
9.54
9.97
8.86
7.89
8.11
8.10
Exporter, Product and Destination
Destination Numbers Tariffs Charged by Partners —All Countries (sd)
16.14
14.87
35.45
15.37
18.86
11.90
12.35
—OECD Countries (mean)
4.91
5.01
5.12
4.98
3.93
4.63
4.75
(sd)
15.58
15.52
14.29
15.56
13.36
12.37
13.34
(mean)
14.86
13.74
14.45
12.43
11.30
10.72
10.42
(sd)
15.12
12.90
46.76
14.28
21.99
10.83
11.04
(mean)
19.16
18.19
13.20
12.01
11.45
9.56
10.02
(sd)
16.52
17.11
9.34
8.16
9.15
5.54
7.29
—Non-OECD Countries
China’s Import Tariffs
12
Table 2: Nominal Exchange Rates, Tariffs and Exports (Quarterly) Full Sample ln(Tariff) ln(NER) Processing Trade
Ordinary Trade Processing Trade ln(Exporter Numbers) -0.0346*** -0.0445** (0.0127) (0.0197) 0.101*** 0.117***
OECD
Non-OECD
-0.00783 (0.0241) 0.0100
-0.0322** (0.0145) 0.0958*** (0.0325) -0.794*** (0.0407) -0.0305*** (0.00960)
-0.0385*** (0.0132) 0.0721***
-0.0384*** (0.0132) 0.0730***
(0.0220) -0.812*** (0.0271)
(0.0219) -0.812*** (0.0271) -0.0475*** (0.00707)
(0.0279)
(0.0201)
-0.0655*** (0.00754)
-0.0987*** (0.0148)
(0.0791) -0.878*** (0.0378) -0.0760*** (0.00853)
1,082,827 0.552
2,100,039 0.595
2,016,485 0.446
ln(China’s Import Tariff) Observations R-squared
4,116,652 0.507
4,116,652 0.507
3,033,655 0.559
ln(Tariff)
-0.0590*** (0.0194) 0.0999** (0.0438)
-0.0588*** (0.0195) 0.101** (0.0437)
ln(Export Value per Exporter) -0.0445** -0.0868*** (0.0186) (0.0302) 0.117** 0.256*** (0.0476) (0.0492)
-0.0232 (0.0357) -0.0404 (0.112)
-0.0349 (0.0241) 0.188*** (0.0582)
-0.372*** (0.0514)
-0.372*** (0.0514) -0.0810*** (0.0138)
-0.113*** (0.0146)
-0.168*** (0.0295)
-0.417*** (0.0714) -0.134*** (0.0158)
-0.445*** (0.0672) -0.0506** (0.0239)
3,033,654 0.408
1,082,827 0.428
2,100,038 0.415
2,016,485 0.342
-0.0412 (0.0409) -0.0304 (0.133)
-0.0345 (0.0234) 0.155** (0.0602) -0.616*** (0.0704) -0.0198 (0.0253)
ln(NER) Processing Trade ln(China’s Import Tariff) Observations R-squared
4,116,651 0.363
4,116,651 0.363
ln(Tariff)
-0.0720*** (0.0219) 0.0867* (0.0440)
-0.0719*** (0.0219) 0.0874** (0.0439)
-0.507*** (0.0479)
-0.507*** (0.0479) -0.0388** (0.0157)
-0.0837*** (0.0161)
-0.151*** (0.0332)
-0.525*** (0.0575) -0.0996*** (0.0179)
4,111,494 0.581
4,111,494 0.581
3,030,170 0.626
1,081,154 0.584
2,096,878 0.596
2,014,488 0.604
Yes Yes Yes Yes
Yes Yes Yes Yes
Yes Yes Yes Yes
Yes Yes Yes Yes
Yes Yes Yes Yes
Yes Yes Yes Yes
ln(NER) Processing Trade ln(China’s Import Tariff) Observations R-squared Product FE Country FE Time FE Cluster by Country
ln(Export Quantity per Exporter) -0.0514** -0.110*** (0.0205) (0.0323) 0.115** 0.252*** (0.0483) (0.0554)
Notes: (1) Since the tariffs charged by Hong Kong are zero, we exclude Hong Kong from the sample. (2) *** p<0.01, ** p<0.05, * p<0.1.
13
Table 3: Nominal Exchange Rates, Tariffs and Exporter Dynamics (Quarterly) Full Sample ln(Tariff)d ln(NER)d Processing Trade
ln(Tariff)d ln(NER)d Processing Trade
Product FE Country FE Time FE Cluster by Country
Non-OECD
-0.0141** (0.00687) 0.00255
-0.0106* (0.00628) 0.384*** (0.123) -0.0541*** (0.00672) -0.00795 (0.00788)
-0.00462 (0.00452) 0.322***
(0.106) -0.0716*** (0.00383)
(0.106) -0.0715*** (0.00384) -0.0199*** (0.00698)
(0.112)
(0.0938)
-0.0236*** (0.00797)
-0.00735 (0.0145)
(0.0980) -0.0794*** (0.00385) -0.0300*** (0.00979)
993,407 0.040
993,407 0.040
713,503 0.048
279,676 0.041
506,763 0.048
486,461 0.042
-0.00736 (0.00893) -0.0456 (0.0748)
-0.00752 (0.00895) -0.0456 (0.0748)
Numbers)d -0.0107** (0.00519) -0.164*** (0.0599)
0.000731 (0.00592) -0.0689 (0.165)
-0.0167 (0.0262) -0.0187 (0.0924)
-0.0940*** (0.00930)
-0.0940*** (0.00930) 0.0120 (0.00871)
0.0218** (0.0105)
-0.0189 (0.0115)
-0.0645*** (0.00480) 0.00983 (0.00991)
-0.127*** (0.0200) 0.0145 (0.0145)
939,768 0.087
939,768 0.087
675,358 0.101
264,165 0.052
481,770 0.050
457,789 0.107
Yes Yes Yes Yes
Yes Yes Yes Yes
Yes Yes Yes Yes
Yes Yes Yes Yes
Yes Yes Yes Yes
Yes Yes Yes Yes
ln(China’s Import Tariff)d Observations R-squared
OECD
-0.00488 (0.00453) 0.321***
ln(China’s Import Tariff)d Observations R-squared
Ordinary Trade Processing Trade ln(New Exporter Numbers)d -0.00359 -0.00778 (0.00511) (0.00516) 0.334*** 0.276***
ln(Exit Exporter -0.00646 (0.0112) -0.00570 (0.0843)
Notes: (1) Since the tariffs charged by Hong Kong are zero, we exclude Hong Kong from the sample. (2) *** p<0.01, ** p<0.05, * p<0.1.
14
Table 4: Real Exchange Rates, Tariffs and Exports (Quarterly) Full Sample
Ordinary Trade Processing Trade ln(Exporter Numbers) -0.0345*** -0.0451** (0.0132) (0.0201)
OECD
Non-OECD
-0.00765 (0.0250)
-0.0333** (0.0151) 0.126** (0.0591) -0.798*** (0.0420) -0.0292***
ln(Tariff)
-0.0390*** (0.0136)
-0.0389*** (0.0137)
ln(RER)
0.139*** (0.0444) -0.812*** (0.0278)
0.139*** (0.0443) -0.812*** (0.0278) -0.0474***
0.140*** (0.0523)
0.179*** (0.0348)
-0.0654***
-0.0971***
0.184* (0.101) -0.873*** (0.0387) -0.0763***
(0.00729)
(0.00785)
(0.0150)
(0.00803)
(0.0100)
1,049,190 0.554
2,047,058 0.598
1,902,943 0.446
-0.0184 (0.0372)
-0.0353 (0.0251) 0.236** (0.0957) -0.437*** (0.0696) -0.0464* (0.0249)
Processing Trade ln(China’s Import Tariff) Observations R-squared
3,950,129 0.508
3,950,129 0.508
2,900,765 0.560
ln(Tariff)
-0.0574*** (0.0202)
-0.0572*** (0.0202)
ln(Export Value per Exporter) -0.0425** -0.0847*** (0.0193) (0.0311)
ln(RER)
0.200*** (0.0707) -0.358*** (0.0514)
0.200*** (0.0707) -0.358*** (0.0514) -0.0804*** (0.0142)
0.190** (0.0823)
0.345*** (0.0710)
-0.112*** (0.0151)
-0.163*** (0.0297)
0.228 (0.138) -0.401*** (0.0699) -0.136*** (0.0153)
2,900,764 0.408
1,049,190 0.428
2,047,057 0.417
1,902,943 0.340
-0.0364 (0.0428) 0.251
-0.0372 (0.0242) 0.207** (0.0925) -0.611*** (0.0732) -0.0202 (0.0262)
Processing Trade ln(China’s Import Tariff) Observations R-squared
3,950,128 0.363
3,950,128 0.363
ln(Tariff)
-0.0715*** (0.0227) 0.193**
-0.0714*** (0.0227) 0.193**
(0.0749) -0.493*** (0.0480)
(0.0748) -0.493*** (0.0480) -0.0404** (0.0161)
(0.0850)
(0.0804)
-0.0856*** (0.0165)
-0.146*** (0.0336)
(0.177) -0.508*** (0.0556) -0.102*** (0.0174)
3,945,113 0.580
3,945,113 0.580
2,897,382 0.626
1,047,557 0.585
2,043,952 0.597
1,901,033 0.605
Yes Yes Yes Yes
Yes Yes Yes Yes
Yes Yes Yes Yes
Yes Yes Yes Yes
Yes Yes Yes Yes
Yes Yes Yes Yes
ln(RER) Processing Trade ln(China’s Import Tariff) Observations R-squared Product FE Country FE Time FE Cluster by Country
ln(Export Quantity per Exporter) -0.0506** -0.108*** (0.0214) (0.0333) 0.183** 0.351***
Notes: (1) Since the tariffs charged by Hong Kong are zero, we exclude Hong Kong from the sample. (2) *** p<0.01, ** p<0.05, * p<0.1.
15
Table 5: Real Exchange Rates, Tariffs and Exporter Dynamics (Quarterly) Full Sample ln(Tariff)d ln(RER)d Processing Trade
ln(Tariff)d ln(RER)d Processing Trade
Product FE Country FE Time FE Cluster by Country
Non-OECD
-0.0108 (0.00741) 0.0446
-0.0120* (0.00615) 0.466*** (0.148) -0.0562*** (0.00683) -0.00892 (0.00825)
-0.00311 (0.00432) 0.373***
(0.123) -0.0729*** (0.00385)
(0.123) -0.0729*** (0.00385) -0.0204*** (0.00717)
(0.128)
(0.114)
-0.0248*** (0.00817)
-0.00599 (0.0150)
(0.0584) -0.0798*** (0.00389) -0.0300*** (0.0101)
954,007 0.041
954,007 0.041
683,307 0.048
270,472 0.041
493,880 0.048
459,945 0.042
-0.00726 (0.00936) -0.148 (0.0953)
-0.00746 (0.00938) -0.148 (0.0953)
Numbers)d -0.0112** (0.00545) -0.241*** (0.0644)
-0.000400 (0.00635) -0.215 (0.144)
-0.0211 (0.0263) -0.102 (0.118)
-0.0943*** (0.00944)
-0.0943*** (0.00944) 0.0142 (0.00884)
0.0237** (0.0107)
-0.0165 (0.0117)
-0.0655*** (0.00482) 0.0107 (0.0101)
-0.127*** (0.0207) 0.0192 (0.0145)
902,532 0.087
902,532 0.087
646,872 0.103
255,414 0.052
469,421 0.050
432,903 0.110
Yes Yes Yes Yes
Yes Yes Yes Yes
Yes Yes Yes Yes
Yes Yes Yes Yes
Yes Yes Yes Yes
Yes Yes Yes Yes
ln(China’s Import Tariff)d Observations R-squared
OECD
-0.00338 (0.00433) 0.373***
ln(China’s Import Tariff)d Observations R-squared
Ordinary Trade Processing Trade ln(New Exporter Numbers)d -0.00196 -0.00680 (0.00486) (0.00535) 0.393*** 0.302***
ln(Exit Exporter -0.00612 (0.0117) -0.112 (0.112)
Notes: (1) Since the tariffs charged by Hong Kong are zero, we exclude Hong Kong from the sample. (2) *** p<0.01, ** p<0.05, * p<0.1.
16
Figure 1: Nominal Effective Exchange Rates China's NEER 130
125
120
115
NEER
110
105
100
95
90
85
80 2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
Note: An increase (decrease) in nominal effective exchange rates means China’s appreciation (depreciation).
17
Figure 2: Tariffs and Nominal Exchange Rates Tariffs by Euro Area and NER Yuan/Euro 2000
2001
2003 Tariffs
2002
NER 2004
Tariffs by USA and NER Yuan/USD
2005
2006
2000
4
12
2001
2003 Tariffs
2002
NER 2004
2005
2006
4.3
8.4
4.1
8.3
3.9
8.2
3.7
8.1
3.5
8
3.3
7.9
3.1
7.8
11 3.5
10
NER
NER
9
Tariffs
Tariffs
3
2.5 8
2 7
1.5
6 2000
2001
2002
2003
2004
2005
2.9
7.7
2006
2000
2001
Tariffs by HKG and NER Yuan/HKD Tariffs 2000
2001
2002
2003
2002
2000
NER
2001
2002
2003
5.5
2004
2003
2004
2005
2006
Tariffs by JPN and NER Yuan/Yen
2005
Tariffs
2004 NER
2005
2006 0.085
2006
1
1.08
0.9
1.07
0.08 5
0.8
1.06 0.075
0.7 1.05 4.5
Tariffs
NER
Tariffs
1.04 0.5 1.03
0.07
NER
0.6
4
0.4 0.065
1.02 0.3 1.01
0.2
3.5 0.06
1
0.1
0
0.99 2000
2001
2002
2003
2004
2005
3
0.055
2006
2000
2001
Tariffs by KOR and NER Yuan/Won Tariffs 2000
2001
2002
2003
2003
2004
2005
2006
Tariffs by GBR and NER Yuan/Pound 2000
NER 2004
2002
2005
2001
2002
2003 Tariffs
4
2006
35
2004 NER
2005
2006 17
0.009
16 0.0085 30
3.5
15
0.008 25 3
14
NER
20
Tariffs
NER
Tariffs
0.0075
13
0.007 2.5 15 0.0065
12
2
10 0.006
5
11
0.0055 2000
2001
2002
2003
2004
2005
1.5
2006
10 2000
2001
2002
2003
2004
2005
Note: A decrease (increase) in nominal exchange rates means Yuan appreciation (depreciation).
18
2006
Figure 3: Tariffs and Nominal Exchange Rates (Continued) Tariffs by SGP and NER Yuan/SGD 2000
2001
2002
2003 Tariffs
0.03
2004 NER
Tariffs by TWN and NER Yuan/TWD 2005
2006
2000 5.2
2001
2002
2003
4
Tariffs
2004 NER
2005
2006 0.28
5.1 0.025
0.27 3.5 5
0.02
4.9
0.26 3
0.015
0.25
NER
Tariffs
NER
Tariffs
4.8
4.7 2.5 0.01
4.6
0.24
4.5 2 0.005
0.23 4.4
0
4.3 2000
2001
2002
2003
2004
2005
1.5
0.22
2006
2000
2001
Tariffs by CAN and NER Yuan/CAD 2000
2001
2002
2003 Tariffs
6.5
2004 NER
2002
2003
2004
2005
2006
Tariffs by IND and NER Yuan/INR
2005
2006
2000 7.5
2001
2002
2003
30
Tariffs
2004 NER
2005
2006 0.195
6 7
0.19 25
5.5 6.5
0.185 20
0.18
4.5 15 5.5
0.175
4
10 5
0.17
3.5
3
4.5 2000
2001
2002
2003
2004
2005
5
2006
0.165 2000
2001
2002
2003
2004
2005
2006
Note: A decrease (increase) in nominal exchange rates means Yuan appreciation (depreciation).
19
NER
Tariffs
6
NER
Tariffs
5
A
Appendix
A.1
Theoretical Motivation
In this appendix, we simply derive implications on the EM from key equations in the Melitz (2003) model. See equations (20) and (19) defining the zero cutoff profit (ZCP) condition in Melitz (2003). These are equations about the relationship between τ , φ∗ , and φ∗x , where τ = 1 + trade costs (tariffs on the home’s exports), φ∗ is the domestic productivity cutoff, and φ∗x is the export productivity cutoff. Then, as shown in Section 8.2 in Melitz (2003), a reduction in τ shifts up the ZCP curve and increases the domestic cutoff φ∗ , which decreases the export cutoff φ∗x and thus increases the EM. Here, let us introduce a nominal exchange rate denoted by e (the price of the foreign currency in the home currency). In the original Melitz model, the export price px = τ ∗ pd = τ /(ρ ∗ φ). In the model with e, on the other hand, px = (τ /e) ∗ pd = (τ /e)/(ρ ∗ φ). So, in the model with e, τ is replaced by (τ /e). Then, under the fixed regime, the argument with (20) and (19) remains unchanged. Under the flexible regime, however, a decrease in τ can be weakened by a decrease in e (domestic appreciation), weakening the effect on the export cutoff φ∗x . Or, a decrease in τ can be strengthened by an increase in e (domestic depreciation), strengthening the effect on the export cutoff φ∗x . Hence, a decrease in τ itself increases the EM, but it can be weakened by a decrease in e (appreciation) or strengthened by an increase in e (depreciation).
A.2
Nominal and Real Exchange Rates
Figures A1 - A2 show the relationship between nominal and real exchange rates for China’s top 10 trade partners from 2000/01 to 2006/12. Note that due to data constraints, to calculate real exchange rates for the euro area we use producer prices for the euro area’s CPI.
A.3
Monthly Results
Tables A1 and A2 show the results of regressions (2.1) and (2.2) at the monthly level.
20
Table A1: Nominal Exchange Rates, Tariffs and Exports (Monthly) Full Sample ln(Tariff) ln(NER) Processing Trade
Ordinary Trade Processing Trade ln(Exporter Numbers) -0.0274** -0.0332 (0.0122) (0.0217) 0.119*** 0.103***
OECD
Non-OECD
-0.00131 (0.0237) 0.00453
-0.0221* (0.0125) 0.0929*** (0.0283) -0.540*** (0.0337) -0.0272*** (0.00998)
-0.0299** (0.0132) 0.0833***
-0.0298** (0.0132) 0.0839***
(0.0214) -0.568*** (0.0228)
(0.0214) -0.568*** (0.0228) -0.0358*** (0.00687)
(0.0289)
(0.0187)
-0.0521*** (0.00693)
-0.105*** (0.0154)
(0.0670) -0.627*** (0.0321) -0.0563*** (0.00838)
2,287,144 0.521
4,716,024 0.544
3,912,478 0.387
ln(China’s Import Tariff) Observations R-squared
8,628,627 0.465
8,628,627 0.465
6,341,341 0.510
ln(Tariff)
-0.0529** (0.0204) 0.119*** (0.0425)
-0.0528** (0.0205) 0.120*** (0.0424)
ln(Export Value per Exporter) -0.0393** -0.0764** (0.0186) (0.0339) 0.149*** 0.239*** (0.0495) (0.0443)
-0.0171 (0.0362) -0.0593 (0.102)
-0.0190 (0.0243) 0.195*** (0.0551)
-0.0483 (0.0559)
-0.0482 (0.0559) -0.0654*** (0.0144)
-0.101*** (0.0141)
-0.182*** (0.0302)
-0.103 (0.0747) -0.109*** (0.0149)
-0.0950 (0.0781) -0.0424 (0.0283)
6,341,340 0.374
2,287,144 0.411
4,716,023 0.382
3,912,478 0.321
-0.0348 (0.0421) -0.0569 (0.131)
-0.0191 (0.0235) 0.154*** (0.0571) -0.263*** (0.0751) -0.0154 (0.0288)
ln(NER) Processing Trade ln(China’s Import Tariff) Observations R-squared
8,628,626 0.343
8,628,626 0.343
ln(Tariff)
-0.0662*** (0.0230) 0.103** (0.0450)
-0.0661*** (0.0230) 0.104** (0.0450)
-0.168*** (0.0486)
-0.168*** (0.0486) -0.0184 (0.0159)
-0.0672*** (0.0149)
-0.175*** (0.0332)
-0.195*** (0.0570) -0.0636*** (0.0176)
8,616,071 0.551
8,616,071 0.551
6,332,627 0.593
2,283,302 0.555
4,708,097 0.559
3,907,849 0.581
Yes Yes Yes Yes
Yes Yes Yes Yes
Yes Yes Yes Yes
Yes Yes Yes Yes
Yes Yes Yes Yes
Yes Yes Yes Yes
ln(NER) Processing Trade ln(China’s Import Tariff) Observations R-squared Product FE Country FE Time FE Cluster by Country
ln(Export Quantity per Exporter) -0.0461** -0.101*** (0.0209) (0.0363) 0.148*** 0.224*** (0.0528) (0.0534)
Notes: (1) Since the tariffs charged by Hong Kong are zero, we exclude Hong Kong from the sample. (2) *** p<0.01, ** p<0.05, * p<0.1.
21
Table A2: Real Exchange Rates, Tariffs and Exports (Monthly) Full Sample ln(Tariff) ln(RER) Processing Trade
Ordinary Trade Processing Trade ln(Exporter Numbers) -0.0268** -0.0334 (0.0125) (0.0220) 0.124*** 0.141***
OECD
Non-OECD
-0.000979 (0.0245) 0.132
-0.0223* (0.0129) 0.104** (0.0475) -0.543*** (0.0346) -0.0263** (0.0104)
-0.0299** (0.0135) 0.120***
-0.0298** (0.0136) 0.120***
(0.0371) -0.568*** (0.0233)
(0.0370) -0.567*** (0.0233) -0.0360*** (0.00700)
(0.0444)
(0.0303)
-0.0517*** (0.00718)
-0.105*** (0.0153)
(0.0904) -0.625*** (0.0328) -0.0572*** (0.00776)
2,228,403 0.524
4,605,377 0.546
3,711,847 0.387
ln(China’s Import Tariff) Observations R-squared
8,317,348 0.466
8,317,348 0.466
6,088,802 0.511
ln(Tariff)
-0.0504** (0.0211) 0.172*** (0.0654)
-0.0502** (0.0211) 0.172*** (0.0654)
ln(Export Value per Exporter) -0.0366* -0.0734** (0.0192) (0.0347) 0.165** 0.285*** (0.0775) (0.0676)
-0.0118 (0.0374) 0.141 (0.136)
-0.0179 (0.0251) 0.210** (0.0861)
-0.0344 (0.0558)
-0.0343 (0.0558) -0.0649*** (0.0148)
-0.0993*** (0.0145)
-0.181*** (0.0300)
-0.0894 (0.0735) -0.112*** (0.0144)
-0.0834 (0.0803) -0.0386 (0.0293)
6,088,801 0.374
2,228,403 0.411
4,605,376 0.383
3,711,847 0.319
-0.0295 (0.0437) 0.159 (0.172)
-0.0202 (0.0242) 0.176** (0.0838) -0.254*** (0.0776) -0.0158 (0.0298)
ln(RER) Processing Trade ln(China’s Import Tariff) Observations R-squared
8,317,347 0.343
8,317,347 0.343
ln(Tariff)
-0.0644*** (0.0237) 0.165** (0.0718)
-0.0644*** (0.0237) 0.165** (0.0718)
-0.155*** (0.0483)
-0.155*** (0.0483) -0.0199 (0.0162)
-0.0678*** (0.0153)
-0.173*** (0.0332)
-0.180*** (0.0552) -0.0669*** (0.0170)
8,305,104 0.550
8,305,104 0.550
6,080,329 0.593
2,224,632 0.556
4,597,585 0.560
3,707,395 0.581
Yes Yes Yes Yes
Yes Yes Yes Yes
Yes Yes Yes Yes
Yes Yes Yes Yes
Yes Yes Yes Yes
Yes Yes Yes Yes
ln(RER) Processing Trade ln(China’s Import Tariff) Observations R-squared Product FE Country FE Time FE Cluster by Country
ln(Export Quantity per Exporter) -0.0443** -0.0986*** (0.0216) (0.0371) 0.162** 0.282*** (0.0820) (0.0828)
Notes: (1) Since the tariffs charged by Hong Kong are zero, we exclude Hong Kong from the sample. (2) *** p<0.01, ** p<0.05, * p<0.1.
22
Figure A1: Nominal and Real Exchange Rates China and Euro Area: NER (Yuan/Euro) and RER 12
NER
China and USA: NER (Yuan/USD) and RER 12
RER
8.4
NER
8.6
RER
8.3 11
11
10
10
8.4
8.2 8.2
8
RER
9
NER
9
RER
NER
8.1
8
8
8
7.8
7.9
7
7
7.6 7.8
6
6 2000
2001
2002
2003
2004
2005
7.7
2006
7.4 2000
2001
China and HKG: NER (Yuan/HKD) and RER 1.08
NER
2002
2003
2004
2005
2006
China and JPN: NER (Yuan/Yen) and RER 1.3
RER
0.085
NER
0.09
RER
1.07 1.25 0.08
0.085
0.075
0.08
0.07
0.075
0.065
0.07
0.06
0.065
1.06 1.2 1.05
1.03
RER
RER
NER
NER
1.15
1.04
1.1
1.02
1.05 1.01
1 1
0.99
0.95 2000
2001
2002
2003
2004
2005
0.055
2006
0.06 2000
2001
China and KOR: NER (Yuan/Won) and RER
2003
2004
2005
2006
China and GBR: NER (Yuan/Pound) and RER 0.009
17
0.0085
0.0085
16
16
0.008
0.008
15
15
0.0075
0.0075
14
14
0.007
0.007
13
13
0.0065
0.0065
12
12
0.006
0.006
11
11
0.0055
10
RER 0.0055 2000
2001
2002
2003
2004
2005
NER
17
RER
2006
RER
RER
NER
NER
NER
0.009
2002
10 2000
2001
2002
2003
2004
2005
2006
Note: A decrease (increase) in nominal/real exchange rates means China’s appreciation (depreciation).
23
Figure A2: Nominal and Real Exchange Rates (Continued) China and SGP: NER (Yuan/SGD) and RER 5.2
NER
China and TWN: NER (Yuan/TWD) and RER 5.2
RER
5.1
0.28
NER
0.29
RER
5.1 0.27
0.28
0.26
0.27
0.25
0.26
0.24
0.25
0.23
0.24
4.9
4.8
4.8
4.7
4.7
4.6
4.6
4.5
4.5
4.4
4.4
RER
NER
4.9
4.3
4.3 2000
2001
2002
2003
2004
2005
0.22
2006
0.23 2000
2001
China and CAN: NER (Yuan/CAD) and RER 7.5
NER
RER
5
NER
5
2002
2003
2004
2005
2006
China and IND: NER (Yuan/INR) and RER 7.5
RER
0.195
NER
0.195
RER
0.19 7
7
0.19
0.185
6.5
6.5
0.185
0.18
RER
NER
6
RER
NER
0.175 6
0.18 0.17
5.5
5.5
0.175
0.165
0.16 5
5
0.17 0.155
4.5
4.5 2000
2001
2002
2003
2004
2005
0.165
2006
0.15 2000
2001
2002
2003
2004
2005
2006
Note: A decrease (increase) in nominal/real exchange rates means China’s appreciation (depreciation).
24
