The environmental effect of trade liberalization: Evidence from China's manufacturing firms

• Author: Jingbo Cui,On Kit Tam,Bei Wang,Yan Zhang
• Published date: 01 December 2020
• Date: 01 December 2020
• DOI: http://doi.org/10.1111/twec.13005

World Econ. 2020;43:3357–3383. wileyonlinelibrary.com/journal/twec

|

3357

© 2020 John Wiley & Sons Ltd

Received: 30 January 2020

|

Revised: 17 June 2020

|

Accepted: 2 July 2020

DOI: 10.1111/twec.13005

SPECIAL ISSUE ARTICLE

The environmental effect of trade liberalization:

Evidence from China's manufacturing firms

JingboCui1

|

On KitTam2

|

BeiWang3

|

YanZhang4

1Division of Social Sciences and Environmental Research Center, Duke Kunshan University, Kunshan, China

2School of Economics, Finance and Marketing, Royal Melbourne Institute of Technology (RMIT) University, Melbourne,

Vic., Australia

3School of Public Administration, University of International Business and Economics, Beijing, China

4School of International Trade and Economics, Central University of Finance and Economics, Beijing, China

Funding information

Jiangsu Qinglan Project; China’s National Natural Science Fund, Grant/Award Number: No. 71603191; Youth Scholar Team

in Social Sciences of Wuhan University “New perspective for development economics research”

KEYWORDS

China, environmental performance, SO2 emissions, trade liberalisation

1

|

INTRODUCTION

Given the increasing concerns over global warming and industrial pollution since the 1990s, discus-

sions on the environmental effects of trade liberalisation have been growing amidst the decades-long

rapid growth in international trade following the creation of the World Trade Organization (WTO;

Copeland & Taylor,2003, 2004). Our paper uses China's WTO accession as a quasi-natural experiment

and our unique firm-level environmental emission database to investigate the relationship between trade

openness and air pollution in China. The accession heralded a swift process of significant trade liber-

alisation that has made China the world's largest producer and exporter of manufactured products. At

the same time, air quality in major Chinese cities has also become the world's most polluted. Thus, an

intensifying but inconclusive debate regarding the effects of trade on the environment has ensued. On

the one hand, as predicted by the pollution haven hypothesis, free trade should worsen China's pollution

problems because firms will relocate their pollution-intensive industries to where the environmental

standards are laxer, or the enforcement of environmental protection is weak (Bombardini & Li,2020;

Taylor, 2005). On the other hand, freer trade could lead to lower pollution through technological effects

via the adoption of pollution abatement technology or technology upgrade (Dean & Lovely, 2010;

Poncet, Hering, & Sousa,2015). We are motivated to investigate the relationship between trade liberal-

isation and pollution with an alternative approach and perspective using a unique firm-level dataset to

produce new evidence on some important areas on which research is rare.

While prior studies have mostly employed macroeconomic data (Antweiler, Copeland, &

Taylor,2001; Cole & Elliott,2003), we aim to explore whether and how trade liberalisation impacts

3358

|

CUI et al.

on the environment at the firm level. Taking advantage of China's accession to the WTO as a qua-

si-natural experiment, we adopt difference-in-differences (DID) methodology to identify the effect of

trade openness with a primary focus on firm-level SO2 emission intensity. Along this line of inquiry,

we disentangle trade liberalisation into input and output tariff cuts for analytical purpose. Moreover,

we are also interested in unveiling whether and how trade liberalisation may impact on the clean-up of

emissions from pollution abatement in the production process or end-of-pipe technologies.

By comparing the SO2 emission intensity of firms in the manufacturing industries that experienced

higher tariff cuts (treatment group) with that of firms in industries subject to lesser tarif f cuts (con-

trol group) during the pre- and post-WTO periods, we document that trade liberalisation leads to a

reduction in SO2 emission intensity at the firm level. Both the pro-competition effects of output tariff

reduction and the input substitution effects from tariff cuts to inputs are found to have contributed to

decreasing SO2 emission intensity. The results are consistent and robust to a series of common trend

and placebo tests, and alternative checks for mitigating concerns about confounding policies.

Our results show that trade liberalisation leads to an improvement in the environmental perfor-

mance of state-owned enterprises (SOE) and private firms, but with a muted impact on foreign-in-

vested enterprises (FIE). The clean-up effect of trade liberalisation on firm-level emission intensity

is found to be more pronounced for firms in cities subject to stringent environmental policy, thus

suggesting there is significant scope for local environmental regulations to improve and reinforce the

positive environment impacts from trade liberalisation. We also find that the two main attributes to

the positive environmental impact of trade liberalisation are the increase in labour resources allocated

to the amount of environmental protection and improvements to the production process that decrease

SO2 emission intensity, rather than the adoption and diffusion of pollution abatement technology.

Since the Uruguay Round negotiations on the General Agreement on Tariffs and Trade in the late

1980s and the implementation of the North American Free Trade Agreement (NAFTA) in the early 1990s,

pioneering studies have centred on the theoretical and empirical relations between trade liberalisation and

environmental pollution. Earlier studies are devoted to the decomposition of the environmental effect

of trade liberalisation into scale, technique and composition effects (Copeland & Taylor, 1994, 1995).

Following this analytical framework, some empirical studies have endeavoured to document evidence

on the environmental impacts of trade at the industry or country level, but the findings have been mostly

inconclusive (Antweiler etal.,2001; Brunel,2017; Cole & Elliott,2003; Cui, Lapan, & Moschini,2016;

Frankel & Rose,2005; Managi, Hibiki, & Tsurumi,2009; McAusland & Millimet,2013).

In this study, we focus on investigating and identifying the firm-level environmental effects

of trade liberalisation. Firm-level emission data have been used in prior studies to estimate

trade effects on firms’ environmental behaviour through within-firm and within-industry adjust-

ments. Considerable firm heterogeneity exists in terms of productivity and managerial ability

in addressing environmental issues, even within narrowly defined industries (Lyubich, Shapiro,

& Walker,2018). Recent studies have investigated the effects of trade liberalisation on firm-

level emissions in the United States (Cherniwchan,2017; Shapiro & Walker,2018) and Mexico

(Gutiérrez & Teshima,2018). Cherniwchan (2017) employs US firm-level emission data to esti-

mate the pollution effects from the NAFTA trade liberalisation. Their findings suggest that trade

liberalisation accounts for about two-thirds of the reduction in air pollution in the US manufactur-

ing sector. Shapiro and Walker (2018) adopt a quantitative structural model to estimate the decline

in pollution emissions in the US manufacturing sector. Their findings show that most of the re-

ductions in air pollutants can be attributed to stringent environmental regulation rather than firm

productivity and international trade. Using Mexican plant-level and satellite data, Gutiérrez and

Teshima (2018) examine the environmental effect of tariff changes due to free trade agreements.

Their findings reveal that import competition contributes to a reduction in plants’ emissions. The