The Impact of Environmental Regulations on Chinese Exports Tongsheng Xu.pdf · The Journal of Global Business Management Volume 12* Number 1 * April 2016 19 The Impact of Environmental

The Journal of Global Business Management Volume 12* Number 1 * April 2016 19

The Impact of Environmental Regulations on Chinese Exports*

Tongsheng Xu1, Yuhua Li1, Hongmei Chen1

School of International Economics and Trade, Jiangxi University of Finance and Economics,

Nanchang, China

ABSTRACT

By employing an extended gravity model, this paper attempts to estimate the impacts of the four

main global environmental regulations - the Kyoto Protocol, the Montreal Protocol, those of the World

Trade Organization and the United Nations Framework Convention on Climate Change (UNFCCC) - and

those of China and its main trading partners, on Chinese exports. The results show that Chinese

environmental regulations and those under the Kyoto and Montreal Protocols and the WTO promote

China’s exports, but UNFCCC hinders exports. China’s trading partners’ environmental regulations

promote exports, but insignificantly. We also conduct various robustness tests.

Keywords: Home Environmental Regulation, Foreign Environmental Regulation, Multilateral

Environmental Regulation, Export, Gravity Model

JEL Code: F13, F18

INTRODUCTION

Environmental regulations directly affect a firm’s production costs and international

competitiveness. However, three different views exist about the relationship between environmental

regulations and exports: Environmental regulations can reduce, increase or not at all affect the volume of

exports (Siebert 1977; Porter, 1991; Grossman & Kruger, 1992)

This paper applies a gravity model to study the impact of environmental regulations on Chinese

exports. In contrast with the existing literature, it has two distinctive features: first, it include four types of

environmental regulations: the Kyoto, Montreal, WTO and UNFCCC, as well as those of China and its

main trading partners; second, it focuses on China’s exports to 24 main trading partners1. The choice to

focus on China is obvious because it is the biggest country in terms of trade volume, and currently the

whole country is exposed to pollution levels unprecedented in human history.

The paper’s main findings are as following:

(1) China’s home environmental regulations significantly promote Chinese exports, and when regulations

are strengthened by 1%, exports increase by 1.6% roughly. Although environmental regulations

increase production costs, they induce firms to innovate, and the innovation effect can outweigh the

cost effect. These may include product innovation, process innovation and environmental technology * Corresponding authors: Tongsheng Xu, School of International Trade and Economics, Jiangxi University of Finance and Economics, Lushan South Road No.169, Nanchang 330013 China. The paper has been financially supported by National Natural Science Foundation of China (71363016, 71263016, and 71303097).

1 In the existing studies, as far as we know, only Li et al. (2012) focus on Chinese exports, and they only include home environmental regulations.

The Journal of Global Business Management Volume 12* Number 1 * April 201620

innovation. Product innovation is aimed at improving environmental protection, product safety, and

heterogeneity. (What do you mean by heterogeneity? Product heterogeneity? If so, say product

heterogeneity.) Process innovation is aimed at increasing the efficiency of resource utilization and

reducing production cost. Environmental technology innovation is aimed at reducing pollution and

saving energy, which not only reduces pollution-control costs (e.g., taxes and fines), but also better

meets international needs.

(2) Trading partners’ environmental regulations promote Chinese exports, but insignificantly. This is

because the contents of partners’ environmental regulations are very similar to China’s, and some are

even more stringent. Thus, these environmental regulations bring similar impacts on Chinese firms so

that the innovation effect outweighs the cost effect, yet the more stringent foreign regulations increase

the difficulty for Chinese exports, making the difference between the two effects not obvious.

(3) The Kyoto, WTO, and Montreal agreements significantly promote Chinese exports. The Kyoto

Protocol requires developed countries to lower annual average emissions by 5.2% during 2008-2012,

compared to 1990, but there is no quota on the reduction of Chinese CO2 emissions, giving Chinese

firms an advantage to increase exports.

The main principle of the WTO is to promote free trade. We find that the effect of trade

liberalization outweighs the cost of environmental regulations. After China’s entry into the WTO, it

revised and formulated many environmental laws and regulations in order to align with WTO regulations,

and to force Chinese firms to innovate.

The effects of the Montreal agreement are similar to the above.

(4) UNFCCC agreements hinder Chinese exports, but insignificantly. This is because the innovation

effect of UNFCCC is slightly less than its cost effect.

(5) While trade partners who signed Kyoto and WTO promote exports, those who signed UNFCC and

Montreal agreements hinder exports.

China is one of the original member states of the United Nations Framework Convention on

Climate Change (UNFCCC, 1992) and the KyotoProtocol (1997) and also is a member of WTO (2001)

and the Montreal Protocol (1991). According to the Environmental Performance Index (EPI) compiled by

Columbia and Yale Universities, China ranked 116th among all 132 participating countries in 2009, which

put it on the same level as Mexico, Vietnam, and Romania, and showed that China's environment is

highly polluted. However, Chinese environmental regulation intensity gradually increased from 43 in

1997 to 67.9 in 2007 (Li, et al., 2010).

Our results have important policy implications. On the one hand, China can benefit from properly

strengthening home environment regulations; on the other hand, Chinese firms should increase R&D

investment and improve environmental technology, to follow home and foreign environment regulations,

especially UNFCCC agreements, in order to reduce greenhouse-gas emission and promote trade.

The remainder of the paper is organized as follows. Section 2 includes a brief literature review.

Section 3 presents the model used, variables, and data. Section 4 conducts the econometric analyses.

Section 5 offers a brief summary and policy implications of this research.

LITERATURE REVIEW

Siebert（1977）argues that environment polices improve firm’s production costs, the countries

with severe environmental regulations do not encourage pollution-intensive production, and the countries

with less stringent environmental regulations increase the comparative advantage of pollution-intensive


production. From the perspective of environment-controlling costs, Carraro & Siniscalco (1992) point out

that severe environmental regulations weaken the competitiveness of a country’s export of products,

especially pollution-intensive products. Beers & Bergh (1997) use a gravity model and a sample of

OECD countries to analyze the impact of environmental regulations on bilateral exports, and find that

severe environmental regulations engender negative and significant impact on the export of OECD

countries. Jug & Mirza (2005) use a gravity model to analyze the impact of environmental regulations in

the EU on its trade, and find that environmental regulations will reduce the exports of both clean and dirty

sector products. likewise. If you meant by the word ‘likewise’ that exports will be reduced equally for the

clean and dirty sectors, use ‘equally’ or ‘roughly the same’ for ‘likewise’. Porter（1991） proposes that environmental regulations first encourage firms to innovate and

improve technology, and then improve their industrial competitiveness. Porter & Linde（1995）further

explain the innovation effect process of environmental regulations, known as the famous “Porter Hypothesis” (Should explain what the ‘Porter Hypothesis’ is.). Frankel（2002）argues that the countries

which develop environmentally-friendly technology innovation have a comparative advantage in

international trade, which is consistent with the Porter Hypothesis. Lu’s (2009) empirical analysis of 95

countries in 2005 based on the Heckscher-Ohlin-Vanek (HOV) model, and argues that a moderate

improvement of environmental regulations gives the countries a comparative advantage on exports of

pollution-intensive products. Li et al. (2012) empirically analyze the impact of China’s environmental

regulations on the comparative advantage of industrial products. The results show that environmental

regulations increase comparative advantage, and the Porter Hypothesis is partially confirmed in China.

Santis (2012) empirically studies the impacts of three multilateral environmental regulations on the

exports of 15 countries in EU, and the results show that the impacts are positive.

Grossman & Kruger (1992) find that the impact of decontamination costs in US industry on trade

patterns between the US and Mexico is insignificant. Wu (2013) uses panel data during 1990-2007 to

analyze the impact of ASEAN countries’ environmental regulations on China’s exports and finds that

when exports are small, it is less affected by environmental regulations; when the proportion of

contaminated products in exports are large, it is largely affected by environmental regulations.

Overall, regarding studies on the impact of environmental regulations on trade, three different

conclusions can be reached due to variances on sample country, time period or econometric method. In

addition, the existing studies mainly focus on the impact of the home environmental regulations in one

country. Few studies address the impact of multilateral environmental regulations on developed countries’

trade. As far as we know, there is no study that combines multilateral environmental regulations with home environmental regulations. Based on Anderson & Wincoop（2003）and Santis（2012）, this

paper adds Chinese home environmental regulations and its trader partners’ environmental regulations to

multilateral environmental regulations in order to empirically test the impact of these regulations on

China’s exports.

MODEL SPECIFICATION AND DATA

The trade gravity model is strongly persuasive in explaining bilateral trade flows and achieves

considerable success in international trade research, so it has been widely employed in the research

on bilateral trade flows. Helpman & Krugman (1986), Evenett & Keller (2002), Anderson & Wincoop

(2003), and Novy(2013) all add to the theoretical foundation of the gravity model.


Model Specification Based on Anderson & Wincoop (2003) and Santis（2012) , this paper extends the gravity model to

estimate the impact of environmental regulations on Chinese exports. First, the explanatory variables of

the standard gravity model are adjusted as follows: by revising the variable GDP and GDP per capita into

the new variable Mass, a measure of import and export countries’ economic scales; by retaining the trade

resistance variable Dist as a proxy of trade costs; by introducing three new multilateral trade resistance

variables -Lan, the national dummy variable for common language, Simil, the similarity index of exports to its trading partner, and actF , factor endowment differences between exports and its trading partner.

Second, six variables of environmental regulations are used, including one domestic environmental

regulation for importers and exporters, respectively, and four major multilateral environmental

agreements (MEAs): Kyoto, UNFCCC, Montreal and WTO, which are all proxies of a dummy variable.

This gives us a new extended gravity model: (1) ijt 0 1 ijt 2 ijt 3 ijt 4 ijt 5 6 it

7 jt 8 ijt 9 ijt 10 ijt 11 ijt ijt

LnExp LnMass LnDist Simil Fact Lan ER

ER Kyoto UNFCCC Montreal WTO

where Ln is the natural logarithm, i is the exporting country, j is the importing country and t is the

year; Expijt is exports in value from country i to country j, that is, the exports of China to its main trading

partners; Massijtis the product of the GDP of the exporting and importing countries, a proxy of economic scale, that is ij i jMass GDP GDP ; Distij is the great-circle distance between two countries’ capital cities

as a proxy for transport costs; Similijtis the similarity index of export to its trading partner’ GDP as a

measure of relative country size:

2ijt it jt it jtSimil Ln 2GDP GDP GDP GDP

Factij, a proxy variable of absolute gap in economic development level between exports and its

trading partner:

ijt it it jt jtFact Ln GDP POP Ln GDP POP

Lan, a dummy variable for common language that equals 1 if the exporting and importing countries have a common language and 0 otherwise; itER , jtER , respectively, are the home and foreign

environmental regulations measured by energy consumption per unit of GDP(GDP/Energy),and is built as:

ER Ln GDP Energy

Kyotoijt, UNFCCCijt, Montrealijt, and WTOijt are dummy variables that equal 1 if both the exporting

and importing countries have signed (and ratified) the Kyoto, UNFCCC and Montreal agreements and 0

otherwise.

In this model, Massijt is a proxy variable for economic scale. The greater the economic scale, the more the bilateral trade flows, so we expect the coefficient 1 has a positive sign; Distij is a proxy variable

of trade costs and is a barrier to trade, so its coefficient 2 is normally expected to be negative; Similijt is

the similarity index of the two trading partners. From the standpoint of inter-industrytrade, the greater the

difference between two countries, the more are the flows of bilateral trade flows, which is similar to the

North-South trade pattern. Likewise, the more similar the two countries, the more are bilateral trade flows. As a result, coefficient for Similijt , 3 , is uncertain. Factij represents the differences in the level of

economic development level. China is the most populous nation with greater differences in GDP per capita from its trading partners, so we expect its coefficient, or 4 , has a positive sign. Lan is a dummy

variable for common language, which shows a stronger relationship between trading partners; when the


two countries have the same or similar language, savings in time and costs of translation will be realized,

total trade costs will decrease considerably, and bilateral trade will increase significantly, so we expect its coefficient 5 has a positive sign as well. itER , jtER measure the intensions of environmental regulations.

China’s strict domestic environmental regulations China may reduce its exports, meaning 6 may be

negative; but strict environmental regulations will also promote the exporter to improve production

equipment and make technical innovations to constantly improve the quality of products, so it can also be positive. Therefore the coefficient for 6 is uncertain. Similarly, 7 , 8 , 9 and 10 are all uncertain. As

for 11 , WTO environmental standards conflict with its basic principal of trade freedom, so 11 is also

uncertain.

Data Sources and Statistical Description

This paper uses 24 countries and regions and 23 years of statistical data from 1990 to 2012. The

countries and regions all have a large volume of trade with China and the statistical data are all relatively

intact. The countries, by region, are: the EU, including the United Kingdom, France, Germany, Italy,

Netherlands, and Belgium; NAFTA: U.S.A, Canada, and Mexico; some countries from the OECD, such

as Japan, South Korea, and Australia; ASEAN, namely Singapore, Malaysia, Thailand, Indonesia,

Vietnam, and Philippines; the transition countries of Brazil, Russia, and India; and other countries, Saudi

Arabia, Hong Kong, and China.

Variables and data sources include Export: http://comtrade.un.org/db/, http://www.chinacustomsstat.

com; GDP: world development indicators (WDI); Population: world development indicators;

Geographical Distance: http://www.cepii.fr.Energy: http://www.wordbank.org. Table A1 shows the

statistical descriptions of dependent variable and independent variables.

In order to intuitively understand the statistical relations between the dependent variables and

independent variables, we make scatter plots to compare trends between different data. Figures 1-6 in the

Appendix present the trends between bilateral exports (LnEXP) and economic scale (LnMass); trade cost

(LnDist); similarity(Simil); differences in factor endowments (Fact); domestic and foreign environmental

regulations(ERit, ERjt); and three multilateral environmental agreements (MEAs)2 during 1990-2012. We find that ln Mass has a significant positive effect on LnEXP, as shown in Figure 1; and LnDist

and Simil have negative effects on LnEXP, as shown in Figures 2, 3 respectively. Due to large differences

in the level of economic development and populations between countries, the trend between LnEXP and

Fact is uncertain, as shown in Figure 4. Compared with foreign environmental regulations (ERjt), China’s

domestic environmental regulations (ERit) have a smaller effect on LnEXP, as shown in Figure 5. And as

for the multilateral environmental agreements (MEAs), Kyoto and UNFCCC agreements have positive

effects on LnEXP, but Montreal has a slight negative effect on LnEXP, as shown in Figure 6.

ECONOMETRIC ANALYSES

LLC Stationary Test

In order to ensure that these panel data can provide effective empirical regression results, we first

make a stationary test. According to econometric theory, a Unit -root test is always used to test if data is stationary, of which the LLC approach（Levin-Lin-Chu) is usually used to test panel data stationary; the

2 Taking into account that the time of entry into WTO of countries in our sample is almost the same, this

paper does not analyze WTO here.


LLC approach is used in this paper. If all the inspection results reject the null hypothesis, that is, no unit

root, then panel data is stationary, otherwise it is non-stationary. Test results are shown in Table 1:

LnEXPijt, LnDistijt, Factijt, ERit, UNFCCCijt, Montrealijt, and WTOijt are all stationary, and LnMassijt, Similijt,

ERjt, Kyotoijt all are first-difference stationary.

Table 1: LLC Stationary Test variables Stats Methods LLC Stats. P value Results

ijtln EXP I -3.47605 0.0003 stationary.

ijtln Mass I&T -9.93494 0.0000 stationary.

ijtln Dist I -3.11495 0.0009 stationary.

ijtSimil I&T -6.69518 0.0000 stationary.

ijtFact N -9.22838 0.0000 stationary.

jtER I&T -6.27834 0.0000 stationary.

jtER I&T -5.30590 0.0000 stationary.

MEAs

ijtKyoto I&T -9.23035 0.0000 stationary.

ijtUNFCCC I&T -3.97399 0.0000 stationary.

ijtMontreal I&T -2.00161 0.0227 stationary.

ijtWTO I&T -3.478302 0.0000 stationary. Notes: (1) In Statistical Methods, I means including intercept term; I&T means including the intercept term and the trend term; N

means both the intercept term and the trend term are not included. The panel data is non-stationary only if the three cannot reject the null hypothesis; Δ means first-order. (2) P > 0.05, accept the null hypothesis, the conclusion is non- stationary; P <0.05, refuse the null hypothesis, the conclusion is stationary.

Results Analyses

As the number of cross-sections (24) is more than the Sequence number (23), we adopt the Statistic

Method of Cross-Section Weights. In order to more clearly analyze the different effects of home, foreign

and multilateral environmental regulations on China’s exports, we split the estimated model into three

different models: Model 1 is testing the effects of home and foreign environmental regulations on China’s

exports; Model 2 is only testing the effects of multilateral environmental regulations on China’s exports;

and Model 3 is testing the effects of all environmental regulations on China’s exports. According to these

three models, we make the fixed effect regression (Within) and random effect regression (GLS),

respectively, by using the software Stata12.0. According to the results of the Hausman-test all three

models should adopt a fixed effect model; the empirical results are as shown in Table 2.

From the estimated results in Table 2, we find that a positive significant export relationship with the

LnMassijt at 1% and a negative significant one with LnDistijt at 1% are confirmed. This is in line with the

relevant empirical literature (Evenett& Keller, 2002; Santis, 2012). It also shows that, the higher the

economic level is, the more trade flows are, and that distance is one of the main factors to hinder exports.

Baltagi et al. (2003) and Santis (2012) show that the more similar two countries are, the more the bilateral

trade flows are, and the sign of Similijt is positive and significant. Yet based on the findings of this paper,

this is positive but insignificant, in contrast to their results. This may be mainly because of the differences

in research objects. Santis (2012) chose as the research object 15 EU countries whose GDP and

population are more similar, which is a kind of the North-North trade pattern, as compared to this paper,

which covers developed and developing countries, resulting in a kind of North-South trade pattern.


Table 2: The Impact of Environmental Regulations on Chinese Export Model 1 Model 2 Model 3

independent variables

FEM REM FEM REM FEM REM

LnMassijt 0.77*** 0.73*** 0.96*** 0.90*** 0.66*** 0.61*** （11.73）（11.95）（18.12）（20.23）（9.26）（8.50）

LnDistijt -0.90*** -0.89*** -0.85*** （-3.63）（-3.46）（-4.17）

Similijt 0.18 0.08 0.23 0.10 0.20 0.12

（1.30）（0.73）（1.55）（0.77）（1.37）（0.92）

Factijt 0.35** 0.31** 0.15 0.11 0.31*** 0.29***

（2.54）（2.25）（1.25）（1.00）（2.84）（2.80）

ERit 1.63*** 1.73*** 1.58*** 1.74*** （4.66）（4.84）（5.07）（5.74）

ERjt 0.24 0.22 0.32 0.25

（0.60）（0.62）（0.76）（0.66）

MEAs

Kyotoijt 0.14* 0.19*** 0.27*** 0.31*** （1.93）（2.67）（3.30）（3.43）

UNFCCCijt 0.34*** 0.37*** 0.07 0.05

（4.50）（4.72）（0.91）（0.71）

Montrealijt 0.72 0.76 0.67 0.69 （1.57）（1.58）（1.57）（1.53）

Language 1.94*** 3.41*** 2.58*** （3.35）（6.14）（4.11）

Constant -22.89*** -12.73*** -31.14*** -20.33*** -17.18*** -7.40*** Hausman-test 27.32*** 40.45*** 27.61***

F-test 174.72*** 151.67*** 119.10*** Notes: (i) ***Significance at 1%; **Significance at 5%; *Significance at 10%. (ii) The data in parentheses are the Statistic p value.

(iii) Because variables LnDistijt and Language show singular matrix problems in the within regression, the respective space is left blank.

The variable Factijt is positive and significant, showing that the greater the differences in economic

development level between importers and exporters, the more helpful it is to increase exports, which

supports the H - O theory. The positive sign and statistical significance of Language proves that language

plays an important role in international trade. When two countries have the same or similar language,

savings in time and costs of translation occur, total trade costs will decrease considerably, and the bilateral

trade will be increased significantly.

As expected, the home environmental regulation variable ERit has a positive and significant impact

on exports, which is different from the conclusions of Beers & Bergh (1997), Grether & Melo (2004),

Mirza (2005). This shows that China’s exports largely benefit from the increasingly strict domestic

environmental regulations, and seems to support Porter’s Hypothesis. Its coefficient is about 1.6 and show

that when regulations are strengthened by 1%, exports increase by roughly 1.6%. The coefficient for

foreign environmental regulations ERjt is positive but insignificant. As for the dummies of four MEAs, in

Model 2, which test the effects of four multilateral environmental regulations on China’s exports, the

coefficients of the four variables all are positive. Kyoto, UNFCCC, and WTO are significant at 5% and

Montreal is insignificant. However, in Model 3, which tests the effects of all environmental regulations on

China’s exports, UNFCCC is negative, though insignificant; Kyoto is positive and significant at 1%;


Montreal and WTO is positive and significant at 10%. There is a difference between conclusions from this

paper and conclusions from Santis (2012). Combining the conclusions of Model 3 with the rapid

development of China’s trade, we can conclude that Kyoto, UNFCCC, and WTO have synergistic effects

on China’s exports, but there is a insignificant conflict effect of UNFCCC on China’s exports. This means

that strict environmental regulations provide incentives for China’s firms to make technical innovations

and increase international competitiveness.

Robustness Test The second column in Table 3 is the last column in Table 2, the third column is robustness test of

sample changes, as Kyoto protocol was not signed during that time, therefore dropped; the fourth column

only considers the robustness of Kyoto regulations on exports among Kyoto, UNFCCC, and Montreal; the

fifth and sixth columns consider only the robustness of UNFCCC and Montreal, respectively; the seventh

column is the robustness test which does not include the ASEAN countries. As shown in Table 3, the

robustness of main conclusions is better.

Table 3: Robustness Test

1990-2012 1990-2000 Kyotoijt UNFCCCijt Montrealijt Non-ASEAN LnMassijt 0.59*** 0.37*** 0.67*** 0.68*** .072*** 0.76*** LnDistijt -0.87*** -0.96*** -1.10*** -1.08*** -1.08*** -82*** Similijt 0.09 0.07 0.06 0.05 0.06 -0.14 Factijt 0.28*** 0.39** 0.37*** 0.34*** 0.45*** 0.80*** ERit 1.59*** 1.35*** 1.88*** 1.89*** 1.76*** 2.07*** ERjt 0.35 0.48 0.28 0.39 0.29 0.02

Kyotoijt 0.32*** dropped 0.08*** 0.08** UNFCCCijt -0.04 -0.02 -0.06** -0.02 Montrealijt 0.77* 0.9* 0.69* 0.28**

WTOijt

0.22* 0.3* 0.003 0.003 0.005 0.04 Language 2.76*** 2.76*** 0.06 0.07 2.76*** Constant -6.07 -5.16 -8.17** -8.92*** -11.03*** -16.22*

Notes: ***Significance at 1%; **Significance at 5%; *Significance at 10%.

To further investigate the impacts of the four MEAs on exports, we include interaction terms between the environmental regulations dummies:

ijtFactyotoK , ijtUNFCCC Fact , ijtMontreal Fact

and ijtWTO Fact . With the inclusions of these terms, the estimated coefficients show the differences in

effects of the variable Factijt on China’s exports between countries (or regions) that had signed MEAs and those that had not. The results are shown in Table 4. We can find that the coefficient of ijtKyoto Fact

is

positive and significant at 10%; the coefficient of ijtFactWTO is positive but insignificant; the coefficients

of ijtFactontrealM and ijtUNFCCC Fact are negative but insignificant. These results show that the

countries or regions that had signed Kyoto and WTO have further positive effects on China’s exports and

are caused by differences in the level of economic development. Nevertheless, the latter’s effect is

insignificant. The countries or regions which had signed UNFCCC and Montreal have a negative effect

on China’s exports, which also are caused by differences in economic development levels, but both of

them are insignificant.


Table 4: Interaction Effects of Dummy variables

Variables Coeff. yoto FactK 0.08*

FactUNFCCC -0.04 ontreal FactM -0.07

FactWTO 0.02

CONCLUSIONS

Using data on China’s exports to 24 main trade partners during 1990-2012, this paper tests the

impact of China’s home environmental regulations, trading partners’ environmental regulations, and

multilateral environmental regulations on China’s exports by using a revised gravity model. The main

conclusions are as follows. (1) China’s home environmental regulations significantly promote China’s

exports, and when regulation strength improves by 1%, exports increase by about 1.6%. This is a synergy

effect. (2) Trading partner environmental regulations promote China’s exports, which is a synergy effect,

though insignificant. (3) Kyoto, Montreal, WTO agreements significantly promote China’s exports, which

is a synergy effect. (4) UNFCCC agreements hinder China’s exports, which is a conflict effect, though

insignificant. (5) Trade partners who signed Kyoto and WTO further promote exports caused by an

economic development gap synergistically, and trade partners who signed UNFCC and Montreal

agreements hinder exports caused by an economic development gap.

These results show obvious policy implications. On the one hand, Chinese companies benefit from

properly strengthened home environmental regulations; on the other hand, Chinese firms also should take

the initiative to increase R&D investment and improve environmental technology and try to adapt to the

home and overseas environmental regulations, especially UNFCCC agreements, in order to make

contributions to both reducing greenhouse-gas emission and trade development.

REFERENCES

Anderson, J. E., Wincoop Van, E. Gravity with Gravitas: A Solution to the Border Puzzle. American Economic Review, (2003). 170-192.

Baltagi, B. H., Egger, P., Pfaffermayr, M. A generalize design for trade flow models. Economic Letters, (2003). 80(3): 391-397.

Beer, V. C., Van Den Bergh, C. J. M. J. An Empirical Multi-country Analysis of the Impact of Environmental Regulations on

Foreign Trade Flows. Kyklos, (1997). 50(1): 29-46.

Carrao, C., Siniscalco, D. Environmental Innovation Policy and International Competition. European Association of Environmental

and Resource Economists, (1992). 2(2): 183-200.

Evenett, S. J., Keller, W. On the Theories Explaining the Success of Gravity Models. Journal of Political Economy, (2002). 110(2):

281-316.

Frankel, J. A. The Environment and Economic Globalization: What is New? Cambridge, (2002).

Grether, J. M.,Melo J. D.“Globalisation and Dirty Industries: Do Pollution Havens Matter?” in R. E. Baldwin and A. Winters (eds.),

Challenges to Globalisation: Analyzing the Economics (Chicago, IL: University of Chicago Press) , (2004). 167-208.

Grossman, M. G., Kruege, B. A. Environmental Impacts of a North American Free Trade Agreement. Discussion papers in

Economics, Woodrow Wilson School, Princeton University, (1992). 644: 158.

Helpman, E., Krugman, P. Market Structure and Foreign Trade. Cambridge, MA: MIT press. (1986).

Jug, J., Mirza, D. Environmental Regulations in Gravity Equations: Evidence from Europe. The World Economy, (2005).

28(11):1591-1615.


Li, G., Ma, Y. & Yao, L. L. The Intensity and Upgrade Path for China’s Industrial Environmental Regulation. China Industrial

Economics, (2010). 3: 31-41. In Chinese.

Li, X. P., Lu, X. X. & Tao, X. Q. Does the Intensity of Environmental Regulation Affect the Competitiveness of Industrial Products

in China? The Journal of World Economy, (2012). (4): 62-78. In Chinese.

Lu, Y. Do Environmental Regulations Influence the Competitiveness of Pollution-intensive products? Economic Research Journal,

(2009). (4): 28-40. In Chinese.

Novy, D. Gravity Redux: Measuring International Trade Costs with Panel Data. Economic Inquiry , (2013). 51(1): 101-121.

Porter, M. A. America’s Green Strategy. Scientifical American, (1991). 168-264.

Porter, M. A, Van der Linde, C. Towards a New Conception of the Environmental Competitiveness Relationship. Journal of

Economics Perspectives, (1995). (09): 97-118.

Santis De, R. Impact of Environmental Regulations on Trade in the Main EU Countries: Conflict or Synergy? The World Economy,

(2012). 35(7): 799-815.

Siebert, H. Environmental Quality and the Gains from Trade. Kyklos, (1977). 30(4): 657-673.

Wu, S. S. The Effect of Environmental Regulation of ASEAN on China’s Export Trade. Communications in Information Science

and Management Engineering（CISME）, (2013). (3): 145-153.

APPENDIX

Appendix 1—Statistical description

Table A1 shows the statistical descriptions of the dependent variable and independent variables.

There is a large difference between maximum and minimum of LnEXP, LnMass, Simil, Fact, LnERi, and

LnERj.

Table A1: Statistical Description Variable Obs. Mean Std. Dev. Min Max LnEXP 552 22.40 1.79 15.17 26.59 LnMass 552 55.02 1.89 49.19 60.12 LnDist 24 8.64 0.70 6.86 9.74 Simil 552 -1.32 0.66 -3.42 -0.69 Fact 552 2.27 1.19 0.05 4.55

Language 24 0.04 0.2 0 1 LnERi 23 0.98 0.28 0.34 1.36 LnERj 552 1.69 0.47 0.26 3.27

MEAs

Kyoto 552 0.33 0.47 0 1 UNFCCC 552 0.81 0.39 0 1 Montreal 552 0.94 0.24 0 1

WTO 552 0.46 0.49 0 1


Appendix 2-- Scatter plots between different data.

Figure 1: Scatter plot between

Exports and Mass Figure 2: Scatter plot between

Exports and Distance Figure 3: Scatter plot between

Exports and Similar

Figure 4: Scatter plot between

Exports and String Figure 5: Scatter plot between Exports and ERit, ERjt

Figure 6: Scatter plot between Exports and Kyoto, Montreal, UNFCCC

The Impact of Environmental Regulations on Chinese Exports Tongsheng Xu.pdf · The Journal of Global Business Management Volume 12* Number 1 * April 2016 19 The Impact of Environmental

Documents