WTO accession, trade expansion, and air pollution: Evidence from China’s county‐level panel data

• Author: Shuai Chen,Faqin Lin,Peng Zhang,Xi Yao
• DOI: http://doi.org/10.1111/roie.12480
• Date: 01 September 2020
• Published date: 01 September 2020

1020

|

wileyonlinelibrary.com/journal/roie Rev Int Econ. 2020;28:1020–1045.

© 2020 John Wiley & Sons Ltd

1

|

INTRODUCTION

A growing debate exists among academics and policymakers about how trade expansion affects the

environment. However, scholars have not reached a consensus due to their heavy reliance on data

from the developed world and data manipulation issue from the developing world.1 In addition, the

evidence of specific mechanisms is missing. This study uses China's, the largest developing country,

data from NASA to fill the research gap, hoping to estimate a reliable impact about trade expansion

on air pollution in the developing world.

In this study, we use Chinese county-level trade and NASA’s air pollution concentration data,

namely, average sulfur dioxide SO2 (µg/m3) and PM2.5 (µg/m3) concentration data. Trade expansion

after World Trade Organization (WTO) accession accounts for approximately 60% and 20% for the

increase of PM2.5 and SO2, respectively, in China. The rising trade-pollution effect is mainly caused by

Received: 2 January 2020

|

Revised: 14 March 2020

|

Accepted: 17 April 2020

DOI: 10.1111/roie.12480

ORIGINAL ARTICLE

WTO accession, trade expansion, and air pollution:

Evidence from China’s county-level panel data

ShuaiChen1

|

FaqinLin2

|

XiYao3

|

PengZhang4

1School of Public Affairs, China Academy

for Rural Development (CARD), Zhejiang

University, Hangzhou, China

2College of Economics and Management,

China Agricultural University, Beijing,

China

3Institute of World Economics and Politics,

Chinese Academy of Social Sciences,

Beijing, China

4School of Management and Economics,

Shenzhen and Shenzhen Finance Institute,

The Chinese University of Hong Kong,

Shenzhen, China

Correspondence

Faqin Lin, College of Economics

and Management, China Agricultural

University, Beijing, China.

Email: linfaqin@cau.edu.cn

Funding information

National Natural Science Foundation of

China, Grant/Award Number: 71773148,

71503281, 71703149

Abstract

This study provides evidence that trade expansion has con-

tributed to the degradation of air pollution in China. On the

basis of different responses of counties’ trade to China's

World Trade Organization accession at the end of 2001,

we exploit air pollution data from NASA to construct a

difference-in-differences predicted trade as an instrument

for our identification. We document statistically significant

and robust evidence on trade expansion, which accounts

for approximately 60% and 20% for the increase of PM2.5

and SO2, respectively, in China. Findings on trade pollu-

tion relation are robust to various tests. Deterioration in the

environment is mainly driven by scale and trade in polluting

sectors.

JEL CLASSIFICATION

F18; F64; O13

|

1021

CHEN Et al.

the size of high-pollution-intensive sectors, which are of first-order importance. Although pollution

intensive trade structure contributes to pollution, it is improving over time.

In addition to reconciling seemingly contradictory results in the literature, we also provide detailed

heterogeneities about the impact of trade expansion on air pollution in China. Moreover, the increas-

ing effect of trade on air pollution is mainly driven by scale and pollution sector intensity, whereas

the technology progress mitigate the impact of trade on air pollution and the pollution sector intensity

is decreasing. These results can enrich our understanding about the impact of trade on pollution and

indicate strong policy implications.

A main issue, often emphasized in the empirical literature, is that trade openness is endogenous in

the regression. First, decisions on whether to trade and how much to trade are clearly not randomly

assigned, wherein regions that trade more may be different from regions that trade less in ways related

to the environment. Second, the regression analysis may be confounded by the feedback going from

environment to trade openness, wherein traders can avoid the polluted regions.

To address such issues, we rely on China's WTO accession as a natural experiment for identifi-

cation. China is a classic example of a country that has undergone rapid development through trade

policies. Given its accession into the WTO, China has grown from a small player in world trade to

the world's largest exporter. At the regional level, China's accession into the WTO has affected some

places more than others as regions differ in their degree of exposure to international trade because of

geography. Coastal regions, for instance, have benefited most from the economic opportunities gener-

ated by China's accession into the WTO. Given that the WTO accession dramatically changed China's

trade pattern by region and time, such an event has therefore been widely used in several previous

studies (Cosar & Fajgelbaum, 2016; Han, Liu, Ural Marchand, & Zhang, 2016; Han, Liu, & Zhang,

2012; Lan & Li, 2015).

Using China's WTO accession as a subject for a quasi-natural experiment, we estimate the effects

of trade openness on air pollution through a difference-in-differences (DID) and instrumental variable

estimation strategy. First, we make use of two sources of sample variation to generate a predicted trade

volume: (a) the difference of trade across counties after China's WTO accession and that of counties

before 2001, and (b) the variation in trade between across counties. These variations enable us to

compare the changes in the trade across counties before and after China's WTO accession in high-ex-

posure versus low-exposure regions and thus estimate the effect of WTO accession on trade. Second,

we use the WTO accession-induced trade as an instrument to run the two-stage least squares (2SLS)

estimation of the effect of trade on air pollution.

This study contributes to three streams of literature. First, our study contributes to the literature by

providing evidence, which can be used to strengthen arguments on whether trade benefits or harms the

environment. On the one hand, trade appears to be good for the environment from some cross-country

analysis (e.g. Antweiler, Copeland, & Taylor, 2001; Copeland & Taylor, 2003, 2004; Frankel & Rose,

2005). These studies utilized data from developed countries. Given that high-income nations have

higher trade and good environmental quality, the regression results often show that trade appears to

be good for the environment. On the other hand, this observation may be overturned to the subset of

less developed countries. Thus, we study the impact of trade on air pollution in the case of the world's

largest developing country.

Second, we use WTO shock as a subject for quasi-natural experiment, which contributes to the

literature utilizing WTO shock to study various topics. For example, China's trade expansion can in-

crease income inequality (Han etal., 2012), productivity (Brandt, Van Biesebroeck, Wang, & Zhang,

2017; Yu, 2015), firm mark-up (Lu & Yu, 2015), expand scope of exports (Feng, Li, & Swenson,

2016), and provide better resource allocation (Feng, Li, & Swenson, 2017; Khandelwal, Schott, &

Wei, 2013) and higher export quality (Fan, Li, & Yeaple, 2015). However, China's trade expansion