Club convergence analysis of regional ecological efficiency in China

• Published date: 01 August 2020
• Author: Zhuo Qiao,Hao Chen
• Date: 01 August 2020
• DOI: http://doi.org/10.1111/1468-0106.12279

ORIGINAL MANUSCRIPT

Club convergence analysis of regional ecological

efficiency in China

Zhuo Qiao

1

| Hao Chen

2,3

1

University of Macau, Taipa, Macau, China

2

Huaibei Normal University, Huaibei, China

3

National Cheng Kung University, Tainan, Taiwan

Correspondence

Zhuo Qiao, Department of Finance and Business

Economics, University of Macau, Avenida da

Universidade, Taipa, Macau, China.

Email: zhuoqiao@umac.mo

Abstract

This paper adopts a network data envelopment analysis

(DEA) model combined with a window analysis approach

to evaluate the ecological efficiency (eco-efficiency) of

30 regions in China over 2000–2015. We also investigate

the existence of eco-efficiency convergence clubs among

these regions and analyse the important factors that drive

eco-efficiency club formation. Our results reveal that,

overall, Chinese regional eco-efficiency deteriorated from

2000 to 2015. We find that there is a significant regional

disparity in eco-efficiency. Our convergence analysis indi-

cates that Chinese regions converge into three eco-

efficiency clubs, suggesting that common economic and

environment policies might have a limited impact on pro-

moting regional eco-efficiency and regionally-tailored

policies need to be designed. Finally, we find that forest

coverage, R&D expenditure and pollution punishment are

important determinants of convergence club membership.

1|INTRODUCTION

China has experienced rapid economic growth since the adoption of its “open-door”economic reform

policy in 1978. Since 2011, it has surpassed Japan and become the second largest economy in the

world after the United States. However, China has been criticized for its development pattern, which

is characterized by low natural resource use efficiency and a high volume of pollution emission. This

pattern of economic development has created considerable pressure on China’s environment and

resources. In this context, steering away from this development pattern, which is detrimental to the

long-term sustainable development of the Chinese economy, has become a key concern of the Chi-

nese Government. Because the assessment of ecological efficiency (eco-efficiency) can provide a sci-

entific basis for this transformation (Mickwitz et al., 2006), eco-efficiency has received much

attention from academic researchers and policy-makers.

The concept of eco-efficiency initially emerged in the 1990s (Schaltegger & Sturm, 1990) and was

subsequently popularized by the World Business Council for Sustainable Development (WBCSD)

Received: 24 May 2017 Revised: 26 March 2018 Accepted: 26 June 2018

DOI: 10.1111/1468-0106.12279

384 © 2018 John Wiley & Sons Australia, Ltd wileyonlinel ibrary.com/journal/paer Pac Econ Rev. 2020;25:384–401.

(Schmidheiny, 1992). It reflects an entity’s ability to produce goods and services while consuming

fewer natural resources and having less impact on the environment (Kuosmanen, 2005; Picazo-Tadeo,

Beltrán-Esteve, & Gómez-Limón, 2012). Data envelopment analysis (DEA) is a widely used efficiency

evaluation approach that can incorporate various inputs and outputs in different dimensions without

definitive weights to aggregate the indicators (Dyckhoff & Allen, 2001; Kuosmanen and Kortelainen,

2005). A number of researchers have adopted this method to analyse the regional eco-efficiency of

China (see Chu, Wu, Zhu, An, & Xiong, 2016; Dai, Guo, & Jiang, 2016; Yang, Jin, Wang, & Lv,

2012; Yin, Wang, An, Yao, & Liang, 2014; Zhang, Bi, Fan, Yuan, & Ge, 2008; Zhang, Liu, Chang, &

Zhang, 2017).

1

The present paper contributes to the emerging literature on Chinese regional eco-efficiency in the

following ways. First, we adopt a network DEA model when evaluating the regional eco-efficiency

in China. This DEA model divides the production system of each region into a production sector and

a pollutant treatment sector and explicitly models how pollutants from the production stage are trans-

formed into desirable outputs in the pollutant treatment stage. Chu et al. (2016) adopt a similar model

to analyse regional eco-efficiency in China in 2013. Their model ignores the fact that some pollutants

remain after the pollutant treatment stage. Our DEA model treats the remaining pollutants as undesir-

able outputs. This new model enables us to evaluate the eco-efficiency more accurately.

When evaluating the eco-efficiency of the Chinese regions using the DEA, studies in the literature

assume that all inputs are transformed into outputs within the same period. However, this assumption

does not fit the actual transformation process because inputs and outputs are distributed over multi-

periods in the real production process. To solve this problem, the present paper adopts a window

analysis approach (Asmild, Paradi, Aggarwall, & Schaffnit, 2004; Charnes, Clarke, Cooper, &

Golany, 1985; Cooper, Seiford, & Tone, 2007). This approach can handle cross-sectional and time-

varying data to measure dynamic effects. This enhances traditional DEA in analysing data over time

because the use of successive and overlapping windows provides a means to assess the temporal

behaviour of decision-making units (DMU). By using this dynamic approach, we can accurately track

eco-efficiency changes of each region through a sequence of time periods.

Finally, we investigate the existence of eco-efficiency convergence clubs among regions in China

using a recent methodology proposed by Phillips and Sul (2007, 2009). We also analyse important

factors that drive club formation. Currently, convergence in eco-efficiency is very important for Chi-

nese policy-makers who are implementing economic and environment policies to achieve country-

wide improvement in eco-efficiency. However, the effectiveness of their policies may have limited

effect if regions are converging to different steady states (i.e. equilibria) by following different con-

verging paths. Therefore, investigating the existence of convergence clubs and analysing the impor-

tant factors that drive club formation can help the government design a regionally-tailored set of

policies targeting each identified club. To the best of our knowledge, our study is the first in the liter-

ature to analyse these issues. The club convergence test we adopt in this study (Phillips & Sul, 2007,

2009) offers a number of interesting features when dealing with Chinese data: (i) we endogenously

determine the number of groups of regions as well as the regions that belong to each club (i.e. this

method uses the information to group the 30 regions into different clubs instead of using predeter-

mined criteria, such as geographic location, to group regions a priori.); and (ii) we explicitly consider

the heterogeneity of our data-generating process across regions and also over time; and (iii) we esti-

mate the speeds of the convergence and the transition paths of eco-efficiency. These results may offer

useful insights for designing effective economic and environment policies.

1

Refer to Section 2 for a detailed review of these papers.

QIAO AND CHEN 385