Carbon Labeling Practice: From the Perspective of Stakeholder’s Interaction (Sustainable Development Goals Series) 9811625824, 9789811625824

Table of contents :
Preface
Contents
1 Carbon Labeling and Its Related Issues
1.1 Introduction
1.2 Carbon Labeling Related Studies
1.3 The Issues Related to Carbon Labeling Practice
References
2 Consumer Behavior Towards Carbon Labeling Scheme
2.1 Introduction
2.2 Carbon Footprint Assessment for Milk Product
2.2.1 Definition of System Boundaries
2.2.2 Source of Inventory Data
2.2.3 Results of Carbon Footprint Assessment
2.2.4 Summary
2.3 A Questionnaire Survey
2.3.1 Design and Collection of Questionnaire
2.3.2 Descriptive Statistic Results
2.3.3 Regression Results
2.3.4 Summary
2.4 A Purchase Decision-Making Experiment
2.4.1 Experimental Design
2.4.2 Experiment Results
2.4.3 Summary
2.5 A System Dynamics Simulation
2.5.1 Model Formulation
2.5.2 Simulation Results
2.5.3 Summary
Appendix
References
3 Interaction Among Stakeholders Involved in Carbon Labeling Scheme
3.1 Introduction
3.2 Fundamentals of Game Theory
3.2.1 Solution by Dominated Strategy Elimination
3.2.2 Solution by Graphic Method
3.2.3 Summary
3.3 A Game Between Consumers and Enterprises
3.3.1 Game Theoretical Model
3.3.2 Theoretical Analysis for Evolutionary Stability
3.3.3 Evolutionary Stability Analysis by Simulation
3.3.4 Impact of Governmental Subsidy Incentives
3.3.5 Summary
3.4 A Game Among Enterprises
3.4.1 Game Theory Based SD Model
3.4.2 An Numerical Case
3.4.3 Simulation Results
3.4.4 Sensitivity Analysis
3.4.5 Summary
3.5 A Game Between Enterprises and Government
3.5.1 SD Model
3.5.2 An Illustrative Case
3.5.3 Simulation Results
3.5.4 Sensitivity Analysis and Discussion
3.5.5 Summary
References
4 Carbon Labeling Improvement and Its Application
4.1 Introduction
4.2 An Improved Carbon Labelling Scheme
4.2.1 Method
4.2.2 Case Examples
4.2.3 Summary
4.3 Carbon Labeling for Benchmarking Emissions Reduction
4.3.1 Method
4.3.2 Results
4.3.3 Summary
4.4 Carbon Labeling for Low Carbon Community
4.4.1 Method
4.4.2 A Case Example
4.4.3 Results
4.4.4 Summary
References
5 Insights and Future Study
5.1 Insights
5.2 Future Study
References
Index

Citation preview

Rui Zhao Yong Geng

Carbon Labeling Practice From the Perspective of Stakeholder’s Interaction

Carbon Labeling Practice

Rui Zhao · Yong Geng

Carbon Labeling Practice From the Perspective of Stakeholder’s Interaction

Rui Zhao Faculty of Geosciences and Environmental Engineering Southwest Jiaotong University Chengdu, China

Yong Geng School of Environmental Science and Engineering Shanghai Jiao Tong University Shanghai, China

ISBN 978-981-16-2582-4 ISBN 978-981-16-2583-1 (eBook) https://doi.org/10.1007/978-981-16-2583-1 © Springer Nature Singapore Pte Ltd. 2021 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd. The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721, Singapore

Preface

In recent decades, global climate change has become a world concern in the social– political issue context. In this regard, carbon labeling is considered as a useful tool since it can reflect the CO2 emission of one product by considering its life cycle. Such a tool may help consumers better understand CO2 emissions of various products so that low-carbon products can be further promoted and consumers can change their lifestyles for pursuing low-carbon consumption. Consumers, government and enterprises are the significant stakeholders in carbon labeling practice. However, their interests have yet to be consistent in promotion of low-carbon consumption. This book focuses on the strategic interactions among possible stakeholders in the carbon labeling practice to explore the opportunities and challenges related to carbon labeling practice, thus to provide insight into low-carbon consumption and production. It is an essential reading for students, researchers, policymakers as well as those with a wider interest in environmental science and sustainable development. Chengdu, China Shanghai, China

Rui Zhao Yong Geng

Acknowledgements This work was financially supported by the National Natural Science Foundation of China (41301639; 72088101; 71810107001; 71690241), the Natural Science Foundation of Sichuan Province for Distinguished Young Scholars (No. 2019JDJQ0020), the Chengdu Science and Technology Project (No. 2020-RK00-00246-ZF) and the Sichuan Province Circular Economy Research Center Fund (No. No. XHJJ-2002; No. XHJJ-2005).

v

Contents

1 Carbon Labeling and Its Related Issues . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 Carbon Labeling Related Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3 The Issues Related to Carbon Labeling Practice . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 1 3 15 16

2 Consumer Behavior Towards Carbon Labeling Scheme . . . . . . . . . . . . 2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Carbon Footprint Assessment for Milk Product . . . . . . . . . . . . . . . . . 2.2.1 Definition of System Boundaries . . . . . . . . . . . . . . . . . . . . . . . 2.2.2 Source of Inventory Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.3 Results of Carbon Footprint Assessment . . . . . . . . . . . . . . . . . 2.2.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 A Questionnaire Survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.1 Design and Collection of Questionnaire . . . . . . . . . . . . . . . . . 2.3.2 Descriptive Statistic Results . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.3 Regression Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4 A Purchase Decision-Making Experiment . . . . . . . . . . . . . . . . . . . . . . 2.4.1 Experimental Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4.2 Experiment Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5 A System Dynamics Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5.1 Model Formulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5.2 Simulation Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

21 21 23 24 26 28 32 32 33 36 37 40 42 43 46 51 52 55 58 66 66 70

3 Interaction Among Stakeholders Involved in Carbon Labeling Scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Fundamentals of Game Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

77 77 79 vii

viii

Contents

3.2.1 Solution by Dominated Strategy Elimination . . . . . . . . . . . . . 3.2.2 Solution by Graphic Method . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3 A Game Between Consumers and Enterprises . . . . . . . . . . . . . . . . . . 3.3.1 Game Theoretical Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.2 Theoretical Analysis for Evolutionary Stability . . . . . . . . . . . 3.3.3 Evolutionary Stability Analysis by Simulation . . . . . . . . . . . . 3.3.4 Impact of Governmental Subsidy Incentives . . . . . . . . . . . . . 3.3.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4 A Game Among Enterprises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.1 Game Theory Based SD Model . . . . . . . . . . . . . . . . . . . . . . . . 3.4.2 An Numerical Case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.3 Simulation Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.4 Sensitivity Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5 A Game Between Enterprises and Government . . . . . . . . . . . . . . . . . 3.5.1 SD Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5.2 An Illustrative Case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5.3 Simulation Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5.4 Sensitivity Analysis and Discussion . . . . . . . . . . . . . . . . . . . . 3.5.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

81 83 85 86 86 89 93 94 97 97 98 101 104 109 109 112 112 115 119 123 126 128

4 Carbon Labeling Improvement and Its Application . . . . . . . . . . . . . . . . 4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 An Improved Carbon Labelling Scheme . . . . . . . . . . . . . . . . . . . . . . . 4.2.1 Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.2 Case Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 Carbon Labeling for Benchmarking Emissions Reduction . . . . . . . . 4.3.1 Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.2 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4 Carbon Labeling for Low Carbon Community . . . . . . . . . . . . . . . . . . 4.4.1 Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4.2 A Case Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

135 135 136 137 139 146 147 148 155 162 163 164 167 167 172 173

5 Insights and Future Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1 Insights . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 Future Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

179 179 183 185

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189

Chapter 1

Carbon Labeling and Its Related Issues

Abstract In the recent decades, global climate change has become a world concern in the social-political issue context. Low carbon economy aims at reduction of greenhouse gas emissions by means of technological innovation, development of alternative energy, industrial upgrading etc. Carbon labeling is considered as an effective measure to raise consumers awareness of climate change, to help change their lifestyles and purchasing behaviors, ultimately to promote low carbon economy. This chapter serves as introductory remarks that provide a brief overview of carbon labeling by using a bibliometric analysis, to address the development of carbon labeling, and the issues related to carbon labeling practice. Keywords Carbon footprint · Carbon label · Carbon labeling scheme · Bibliometric analysis

1.1 Introduction With the issue of global climate change assuming a higher profile in the socio-political context, greater attention is being paid to the greenhouse gas emissions produced by goods and services, directly and indirectly during their lifecycle (Wiedmann and Minx 2008; Iribarren et al. 2010). Business management, in the UK and elsewhere, has been encouraged, and in some instances required, to measure and report carbon emissions to incentivize and monitor emissions reductions (e.g., DEFRA 2009). For manufactured goods environmental interventions, such as emissions reductions, need to be implemented mainly through product design and innovation (Deutz et al. 2010; Song and Lee 2010). Carbon footprinting is considered as an effective measure to ensure the emissionreduction targets are met, thus to promote the development of the low carbon economy (Carbon Trust 2008a; Gomi et al. 2010). The carbon footprint becomes a significant indicator of environmental impact, interpreted as the total amount of carbon dioxide, or its equivalent to greenhouse gas emissions, which is directly or indirectly emitted by a specific activity, e.g., a product, service etc. (Carbon Trust 2008a; Wiedmann and Minx 2008; Röös et al. 2013; Fang et al. 2014). On such basis, carbon labeling is a tag summary to provide the carbon footprint information for consumers, in order © Springer Nature Singapore Pte Ltd. 2021 R. Zhao and Y. Geng, Carbon Labeling Practice, https://doi.org/10.1007/978-981-16-2583-1_1

1

2

1 Carbon Labeling and Its Related Issues

to enhance consumers awareness related to sustainable consumption (Brenton et al. 2009; Tan et al. 2014). Carbon label is actually a family member of eco-labels, in which it focuses on a single category of environmental impact, i.e., climate change, whilst the eco-labels provide information to reflect environmental superiority of awarded products, by measuring a number of environmental impacts, including climate change, ozone depletion, waste disposal, acid deposition, etc. (Wu et al. 2014a, 2015). It has been proposed by UK Government and the Carbon Trust in 2006, with 2 years of the labeling validity (Guenther et al. 2012). Since then, many developed countries initiated their carbon labeling systems by considering their own contexts, including US, Canada, Japan etc (Liu et al. 2016). The information provided by the carbon labeling may be presented in a form of either numerical value or reduction commitment of embodied product (Wu et al. 2015). The carbon labeling is seen as an effective means of communication to raise consumers awareness of climate change, and to help change their lifestyles and purchasing behaviours (Carbon Trust 2008b; Tan 2009). More than 90 brands from 19 countries have conducted carbon labeling certification, including Tesco, Dyson, and Kingsmill (Zhao and Zhong 2015). Consumers are open market drivers, and their acceptance of carbon labelled products is an important factor in influencing whether enterprises are willing to attempt carbon labelling accreditation. Related study has shown that 59% of consumers select at least one carbon labelled product in shopping, which reveals that consumers’ enthusiasm for purchasing carbon labelled products is gradually increasing (Messum 2012). However, the actual demand still remains far from those expected by enterprises (Gupta and Ogden 2009; Gleim et al. 2013). The key issue lies in the price premium, which indicates the higher price paid by consumers for carbon labelled products, in order to receive their environmental quality (Milovantseva 2016). However, consumers tend only to accept low product premiums, because their willingness to pay for carbon labelled products are significantly affected by individual characteristics such as age, gender, income, education level etc. (Ramayah et al. 2010; Olive et al. 2011). For enterprises, they have to face market risk caused by the product premiums as well, which may give rise to uncertainty on their long term commercial success (Tian et al. 2014). This is especially true for the exiting carbon labelling systems, which are generally optional for enterprises to implement (Shi 2013; Tan et al. 2014). They lack sufficient driving force from external policy, thus may reduce their willingness to have carbon labelling certification (Zhao et al. 2017). Governments play a leading role in developing well-designed policies to drive industrial innovation into product sustainability and thus promote sustainable performance (Kanada et al. 2013; Choi 2015). Sustainable performance requires coordination among all participating agents to work together to create a win–win outcome (Choi 2015). However, it is difficult to visualize the performance of these policies, due to the complexity of a sustainable operation for all the participants, e.g., trustworthy, loyal partnership etc. (Myeong et al. 2014; Choi 2014, 2015). Additionally, inflexible policymaking may be ineffective if enterprises do not respond actively (Kane 2010).

1.1 Introduction

3

From above analysis, it is obvious that the implementation of a carbon reduction labeling scheme involves the coordination of various stakeholders such as governments, enterprises, and consumers. Specifically, benefits to enterprises, to consumers and to government have yet to be consistent in the current market of carbon labelled products. How to predict their strategic interactions has a practical significance to accelerate low carbon consumption. Since there is no carbon-labeling scheme existed in China, this book identifies the influence in developing a carbon-labeled product market, to encourage the involved stakeholders to promote low-carbon consumption and production. This chapter introduces the development of carbon labelling related studies. Chapter 2 investigates consumers’ perception to carbon-labeling scheme by using empirical studies and a system dynamics (SD) simulation. Chapter 3 applies game theory to simulating strategic interactions among enterprises, consumers and government in a carbon-labeled product market. Chapter 4 presents the improvement of carbon labeling and its application to products and services. The last chapter summarizes the research work, and provides insight into policy implications to promote carbon labelling development.

1.2 Carbon Labeling Related Studies This section provides a bibliometric analysis to review the research progress of carbon labeling scheme during the period of 2007–2019, to provide insight into its future development. The number of publications, the countries of publications, the authors, the institutions, and the highly cited paper were included for statistical analysis. CiteSpace software package was used to visualize the national collaboration, keywords co-appearance and aggregation. The literature data were obtained by a predefined information retrieval from the Web of Science (WOS) Core Collection database, specifically from the sub-databases of Science Citation Index Expanded, Social Sciences Citation Index and Arts and Humanities Citation Index. The predefined entries were set based upon the “carbon label*” and the highlighted words in its associated definition, e.g., environmental impact, life cycle assessment, product, service etc. There were 2016 records at the retrieval time 14/5/2020, and their document types were confined to Article or Review in English, as shown in Table 1.1. However, most of the publications focused on analytical chemistry, physical chemistry, biochemistry molecular biology, microbiology etc., which were apparently not conform to the research topic. After removal of the duplicated and irrelevant publications, there were 175 articles finally identified for further bibliometric analysis. The retrieval results were performed by descriptive statistics corresponding to a set of bibliometric indicators, such as number of publications, country, category, journal, institution, authors, highly cited papers and keywords, to investigate the attentions to the carbon label related studies. Specifically, Microsoft Office Excel and bibliometric online analysis platform were applied to implementing the statistical analysis. The

4

1 Carbon Labeling and Its Related Issues

Table 1.1 Criteria for data retrieval Set

Results

Search criteria

#6

2016

(#5 OR #4 OR #3 OR #2 OR #1) AND LANGUAGE: (English) AND DOCUMENT TYPES: (Article OR review) Indexes = SCI-EXPANDED, SSCI, A&HCI TIMESPAN = 2007–2019

#5

210

(TS = (carbon label*) AND TS = (environmental impact)) AND LANGUAGE: (English) AND DOCUMENT TYPES: (Article OR review) Indexes = SCI-EXPANDED, SSCI, A&HCI TIMESPAN = 2007–2019

#4

378

(TS = (carbon label*) AND TS = (consume*)) AND LANGUAGE: (English) AND DOCUMENT TYPES: (Article OR review) Indexes = SCI-EXPANDED, SSCI, A&HCI TIMESPAN = 2007–2019

#3

68

(TS = (carbon label*) AND TS = (life cycle assessment)) AND LANGUAGE: (English) AND DOCUMENT TYPES: (Article OR review) Indexes = SCI-EXPANDED, SSCI, A&HCI TIMESPAN = 2007–2019

#2

68

(TS = (carbon label*) AND TS = (service)) AND LANGUAGE: (English) AND DOCUMENT TYPES: (Article OR review) Indexes = SCI-EXPANDED, SSCI, A&HCI TIMESPAN = 2007–2019

#1

1601

(TS = (carbon label*) AND TS = (product)) AND LANGUAGE: (English) AND DOCUMENT TYPES: (Article OR review) Indexes = SCI-EXPANDED, SSCI, A&HCI TIMESPAN = 2007–2019

values corresponding to journal impact factors were from Journal Citation Reports in 2018. CiteSpace software package was used to visualize the underlying connections among keywords by generation of a co-occurrence network. Number of Publications As shown in Fig. 1.1, there were 175 publications in the period of 2007–2009 with a gradual increase. Particularly, there was a peak of publications in 2016. Less than five papers published annually during the period of 2007–2009, which highlighted the infancy of carbon labelling scheme, as it was issued by UK in 2006. From 2009 to 2011, the number of publications increased rapidly. During 2011–2013, it had remained unchanged, as 10 publications annually. Form the year of 2014, the growth was significant towards a peak in 2016, which had an increase of 15.5 times of the publications compared with that in 2007. A possible reason might be that Paris Climate Agreement drives structural transformation of the global carbon markets, which calls for effective market-based policy tools, e.g. carbon labeling scheme, eco-labeling scheme, carbon trade mechanism, to promote emissions reduction and energy transformation (Aldy et al. 2016; Fujimori et al. 2016; Falkner 2016). Since 2017, the number of publications had decreased 65% and remained a slow growth till 2019. Countries of Publications There were 43 countries and regions have contributed articles in carbon labeling related research areas. Table 1.2 shows the top 10 countries in terms of their publications during 2007–2019. It is apparent that U.S. ranks first, China as second, followed

1.2 Carbon Labeling Related Studies

5

Fig. 1.1 Number of publications during 2007–2019

by UK, Germany, Australia, and Italy. In particular, EU countries contributed the most of publications, indicating a regional spillover effect due to their close collaborations in economy and politics. Table 1.2 Top 10 countries with the most published articles from 2007 to 2019

Country

Centrality

TP

Percentage (%)

USA

0.98

41

23.43

China

0.21

27

15.43

UK

0.38

23

13.14

Germany

0.19

16

9.14

Australia

0.19

14

8.00

Italy

0.92

13

7.43

Netherland

0.70

10

5.71

Sweden

0.08

10

5.71

France

0.61

9

5.14

South Korea

0.82

7

4.00

TP Total publications

6

1 Carbon Labeling and Its Related Issues

Figure 1.2 shows the variation of number of publications regarding the top 5 countries, where there is a peak occurred between 2015 and 2018. United Kingdom and United States are the pioneer of carbon labeling scheme and have been involved in such studies for 12 years, indicating a remarkable spillover effect in policy making on low-carbon consumption and production. Carbon label related studies were still in progress in developing entities, e.g., China engages on such issue until 2012. However, China makes great contribution in low-carbon development due to great pressure on energy structure (Liu et al. 2012). In such context, China calls for such a labeling system to change the consumption patterns while upgrading the supply chain to further improve product/service quality (Wang et al. 2015). The national collaboration provides a new perspective to assess the academic impact of countries in carbon label related studies. Figure 1.3 shows the network of national collaboration, in which each country is denoted by a circle. The size of the circle represents the frequency of collaboration. The higher the frequency, the larger the circle. The thickness of the purple circle represents the centrality. The higher the centrality, the thicker the purple circle. The centrality is used to indicate the international status of a country in a certain research area. The US has the largest degree of centrality (0.98), implying a comparative high international influence in this research field. Italy, in spite of only 13 articles being published, its centrality ranks the second (0.92), indicating great potentials in development of

Fig. 1.2 Number of publications regard the top 5 countries

1.2 Carbon Labeling Related Studies

7

Fig. 1.3 National cooperation network

international cooperation. For instance, at least 2 authors were from other countries in the 13 published papers. China, whilst ranked as the second in the number of publications, its centrality was relatively low, which further calls for reinforcement of international cooperation to share the knowledge. Subject Categories and Published Journals There were wide topics involved in the studies of carbon labeling scheme during the period of 2007–2019, which were subject to 44 categories. Figure 1.4 shows the top 10 categories, and the most common category was “Environmental Sciences”, accounting for 23.1%. The subject categories were mainly related to environment,

Fig. 1.4 Top 10 subject categories

8

1 Carbon Labeling and Its Related Issues

economy, food and business, which demonstrated that carbon labeling related studies have a multidisciplinary field. Carbon label related studies were published in 85 journals. Table 1.3 shows the top 10 journals, where 86 articles were published, accounting for 49.14%. Most of the journals were from Elsevier publisher. The Journal of Cleaner Production is highlighted by both of the maximum publications and the total citations. The Journal of Renewable Sustainable Energy Reviews has the highest impact factor, and the Journal of Food Policy has the highest average number of citations. Distribution of Research Institutions Table 1.4 shows the top 5 institutions sorted by the number of first author’s publications. There were 287 institutions involved in carbon label related studies during 2007–2019. Southwest Jiaotong University was identified as the institution where the most of publications contributed, followed by Central Queensland University, Swedish University of Agricultural Sciences, Eidgenössische Technische Hochschule Zürich (ETH) and Korea University. Southwest Jiaotong University published 6 papers in the area, where it mainly focused on application of system Table 1.3 Top 10 journals with the most published articles Journals

IF (2018) EP Percentage (%) TC

ACP

Journal of Cleaner Production

6.395

37

21.14

149 4.03

Food Policy

3.788

10

5.71

91 9.10

Sustainability

2.592

10

5.71

1 0.10

Energy Policy

4.880

6

3.43

3 0.50

International Journal of Life Cycle Assessment

4.868

5

2.86

7 1.40

British Food Journal

1.717

4

2.29

24 6.00

Appetite

3.501

4

2.29

19 4.75

10.556

4

2.29

14 3.50

Environmental Science & Policy

4.816

3

1.71

26 8.67

Energy Economics

4.151

3

1.71

21 7.00

Renewable & Sustainable Energy Reviews

EP Entire publications, TC Total citations, ACP Average citation per paper

Table 1.4 Institutions with the most publications Institution

Country

Southwest Jiaotong University

China

6

Central Queensland University

Australia

4

Swedish University of Agricultural Sciences

Sweden

4

Eidgenössische Technische Hochschule Zürich

Switzerland

4

Korea University

Korea

3

TP Total publications

TP

1.2 Carbon Labeling Related Studies

9

dynamics and game theory to investigate the interaction among consumer, enterprises and government in implementation of carbon reduction labeling scheme, to provide policy implications on sustainable consumption and production (Zhao et al. 2016, 2018a). Central Queensland University proposed carbon labeling as an indicator to reflect the environmental impact of building materials in order to achieve green architecture design (Wu and Feng 2012; Wu et al. 2014b). Swedish University of Agricultural Sciences was ranked as the fourth, where they paid close attentions to measurement of uncertainties in food products carbon footprinting (Röös et al. 2010, 2011). ETH and Korea University gave emphasis on consumers’ preferences and willingness to pay for carbon labelled products, by which the possible influencing factors were explored (Van Loo et al. 2014; Lazzarini et al. 2017; Shi et al. 2018). Highly Cited Papers The top 10 highly cited papers in the field of carbon labeling scheme during the period of 2007–2019 are given in Table 1.5. These papers were distributed in 7 journals, and three of the highly cited papers were published on Journal Food Policy. According to the statistics of the institution where the first author was located, the article published by Grunert et al. (2014) ranked the first with 267 citations. In these highly cited papers, some of which focused on comparing utilities of various sustainable labeling scheme, including carbon label, environmental and ethical labels (Van Loo et al. 2014; Lazzarini et al. 2017; Shi et al. 2018; Onozaka and McFadden 2011; Grunert et al. 2014). Most of the citations related to these highly cited papers are mainly from methodological perspectives to focuse on public attitudes towards organic foods, by means of choice experiments, structured interviews, and questionnaire survey (Vermeir and Verbeke 2006; Janssen and Hamm 2012). The co-citations are highlighted by exploration of consumers’ preference and willingness to pay for organic foods by using Schwartz’ values theory combined with the planned behavior theory (Aertsens et al. 2009; Zander and Hamm 2010). However, several studies firmly believed that carbon label may provide a signal to help consumers change their purchasing behaviors if they can fully understand the associated labeling information (Peschel et al. 2016; Camilleri et al. 2019). In such context, Rugani et al. (2013) argued if the underlying method for carbon footprinting, i.e., the life cycle assessment, can be improved as more transparency. In summary, these highly cited papers had addressed the challenges in application of carbon labelling scheme, to lay out foundation for its future development. Keywords Keywords can be used to reflect the hotspots and topics of research interest in a certain time period (Yang and Meng 2019). CiteSpace software package was employed to produce a keywords co-occurrence network (Chen and Song 2019). In such process, a number of synonymous keywords were sorted by merging similar words, such as “Carbon label” and “Carbon labeling”, “greenhouse gas” and “GHG”, “life cycle assessment” and “LCA”, etc. Table 1.6 shows the frequency regarding the keywords occurred during 2007 to 2019. There were 98 keywords obtained, among which 22

10

1 Carbon Labeling and Its Related Issues

Table 1.5 Top 10 cited papers Title

Author

Country and institution

Journal and year

TC

ACP

Sustainability Grunert, K. G.; labels on food Hieke, S; Wills, J products: consumer motivation, understanding and use

Denmark, Aarhus Food Policy (2014) 267 University

44.50

Does local labelling complement or compete with other sustainable labels? A conjoint analysis of direct and joint values for fresh produce claim

Norway, University Stavanger

American Journal of Agricultural Economics (2011)

144

16.00

England, Northumbria University

Food Policy (2011) 112

12.44

Onozaka, Y.; McFadden, D. T

The use and Gadema, Z.; usefulness of Oglethorpe, D carbon labelling food: a policy perspective from a survey of UK supermarket shoppers Carbon labelling of grocery products: public perceptions and potential emissions reductions

Upham, P.; England, Dendler, L.; Bleda, University M Manchester

Journal of Cleaner Production (2011)

108

12.00

Consumers’ valuation of sustainability labels on meat

Van Loo, E. J.; South Korea, Caputo, V.; Nayga, Korea University R. M.; Verbeke, W

Food Policy (2014)

85

14.17

International Journal of Physical Distribution & Logistics Management (2010)

68

6.80

Product-level McKinnon, A. C carbon auditing of supply chains environmental imperative or wasteful distraction?

Scotland, Heriot Watt University

(continued)

1.2 Carbon Labeling Related Studies

11

Table 1.5 (continued) Title

Author

Country and institution

Journal and year

TC

ACP

The potential role Cohen, M. A.; USA, Vanderbilt of carbon labelling Vandenbergh, M. P University in a green economy

Energy Economics (2012)

67

8.38

Finnish consumer perceptions of carbon footprints and carbon labelling of food products

Hartikainen, H.; Finland, MTT Roininen, T.; Agrifood Katajajuuri, K. M.; Research Finland Pulkkinen, H.;

Journal of Cleaner Production (2014)

64

10.67

Vulnerability of exporting nations to the development of a carbon label in the United Kingdom

Edwards-Jones, G.; Wales, Bangor Plassmann, K.; University York, E. H.; Hounsome, B.; Jones, D. L.; Canals, L

Environmental Science & Policy (2009)

64

5.82

British Food Journal (2011)

51

5.67

Challenges of Roos, E.; carbon labelling of Tjarnemo, H.; food products: a consumer research perspective

Sweden, Swedish University of Agricultural Sciences

TC Total citations, ACP Average citation per paper

keywords appeared above 10 times. During the time period of retrieval, “carbon footprint,” “willingness to pay,” and “food,” were the top 3 keywords, indicating that surveys on consumer’s attitudes towards carbon labelled product have aroused wide academic concerns in the past 12 years. Figure 1.5 shows the keywords co-occurrence network. The size of the circle represents the occurrence frequency. The lines between the nodes denote their connections. The thicker of the lines, the stronger the connection. The lines between nodes are bright in color, indicating that there are a number of research hotspots derived in recent years. The largest circle is “carbon footprint”, by which “carbon label,” “willingness to pay,” “food,” and “attitude” are closed linked. The node with the highest centrality is “carbon label”, by which there are six nodes connected, including “carbon footprint,” “information,” “life cycle assessment,” “performance” etc. Such phenomenon may imply that life cycle-based assessment is fundamental to the performance of carbon labeling scheme, through which various form of carbon footprint information is provided. Besides, consumer behavior towards carbon labelling scheme is full of the academic research interests. Consumers are receptors of carbon labelled products or services, and their purchase intentions are critical to implementation of the labeling policy (Shuai et al. 2014; Zhao et al. 2018b).

12 Table 1.6 Descriptive statistics of the keywords

1 Carbon Labeling and Its Related Issues Keywords

Count

Centrality

Percentage (%)

Carbon footprint

50

0.18

7.72

Willingness to pay 37

0.18

5.38

Food

28

0.03

4.07

Carbon label

26

0.38

3.78

Consumption

26

0.20

3.78

Product

24

0.07

3.49

Climate change

21

0.10

3.05

Information

20

0.32

2.91

Behaviour

18

0.13

2.62

Greenhouse gas emission

18

0.04

2.62

Life cycle assessment

17

0.19

2.47

Eco label

16

0.21

2.33

Sustainability

16

0.23

2.33

Impact

15

0.06

2.18

Choice

14

0.12

2.03

Footprint

14

0.12

2.03

Perception

14

0.05

2.03

Attitude

13

0.20

1.89

Label

13

0.03

1.89

Choice experiment

12

0.08

1.74

Preference

11

0.10

1.60

Policy

10

0.14

1.45

Keywords clustering was conducted to identify the research frontiers of carbon label related studies. There were 7 clusters identified, and their corresponding silhouette values were above 0.7 (if above 0.5, the cluster is considered to be reasonable), as shown in Fig. 1.6. The largest cluster (#0) was entitled “categorisation task”, which put emphasis on application of classification to ensure reliable comparisons among similar products with different carbon emissions. The second cluster (#1), was given as “uncertainty analysis”, focusing on uncertainty regarding carbon footprint assessment. The third cluster (#2), as “carbon footprint label”, mainly paid attention to its relative performance, e.g., the environmental impact, energy efficiency etc. The cluster (#3), as “climate change”, indicating public awareness of the implication of carbon labeling scheme. Burst detection is used to determine if any change occurred at the research hotspots (Chen 2006). The bursting words refer to the keywords that have suddenly emerged or increased significantly in a short period of time, which may provide insight into

1.2 Carbon Labeling Related Studies

13

Fig. 1.5 Keywords co-occurrence network

Fig. 1.6 Clusters of carbon label studies related keywords

identification for future research interests (Zhu and Hua 2017). It is detected by an algorithm proposed by Kleinberg, which generates a list of important words in terms of their frequencies in a finite duration of time (Chen 2006). The frequency of the word changes implies possible state transition, indicating as a burst (Onozaka et al. 2016). In Fig. 1.7, the red rectangle indicates the strongest bursts, since the corresponding keywords have multiple occurrences in the specific time nodes. “Strength” shown in Fig. 1.7, indicates the bursting words that have been mentioned frequently than any other words in a specific time period (Chen 2017). There are 15 keywords with apparent bursts in this study. Such phenomenon implies that the carbon labeling

14

1 Carbon Labeling and Its Related Issues

Fig. 1.7 Top 15 keywords ranked by burst detection

studies have been distinguished by three stages: first from 2007 to 2012, as the carbon labeling scheme was in its infancy, and the burst keywords mainly contained carbon footprint, carbon label, carbon emission, food and energy (Tan et al. 2012). Especially, the labeling policy was gradually extended to the household equipment and construction industry to evaluate its energy efficiency. The second stage was from 2013 to 2015, where the keywords were booming, with life cycle assessment, eco label, food consumption and market covered. In this stage, studies preferred the utility of the labeling policy and its possible impact on production, trade and export. For instance, carbon labeling scheme was compared with other eco-labels to highlight its impact on the development of trade and economy (Röös and Tjarnemo 2011). In the third stage (2016–2019), studies placed emphasis on individuals’ behavior towards carbon labelled product or service by conducting surveys to investigate their perception, preference and willingness to pay (Spaargaren et al. 2013; Zhao and Zhong 2015; Zhao et al. 2017; Chen et al. 2018). At the same time, multi-stakeholders’ interaction was gradually involved in the carbon labeling relevant researches (Juvan and Dolnicar 2016). The application of carbon labelling scheme was gradually transformed from product to service, e.g. tourism. Carbon

1.2 Carbon Labeling Related Studies

15

labelling may affect the behaviour of tourist who purchases low-carbon tourism services, even though tourist may have pro-environmental intent or not (Hu et al. 2019).

1.3 The Issues Related to Carbon Labeling Practice Though the analytic review on the 175 carbon label related articles, three main challenges have been identified to lay out foundations for future studies. First, precise accounting for carbon emissions is a prerequisite for the labeling practice (Röös et al. 2010). Carbon label is also entitled carbon footprint label, since most of the carbon labels are presented in a footprint form (Liu et al. 2016). As implied in the definition of carbon label, lifecycle-based carbon footprinting is a cornerstone to support the presentation of carbon labeling (Cohen and Vandenbergh 2012). However, the system boundary for a specific product or service is difficult to define, which may cause uncertainty in expression of the accounting results, thus decreasing credibility of the labeling scheme (Wu et al. 2014b). For instance, existing studies have shown that the carbon footprint of crops are varied by place of production, due to the different system boundary for lifecycle accounting, giving rise to uncertainty in food carbon labeling practice (Röös et al. 2011). Such inconsistency may give rise to the same product that has significantly different numerical values labelled on its package (Zhao et al. 2012). It thus calls for improvement of carbon footprint assessment to ensure fair comparison among similar products or services. Besides, a functional unit is generally followed by the life cycle assessment, which limits comparability among various types of products (Röös et al. 2010). There is a call for standardized methods to normalize carbon footprint regarding product or service into a common scale, in order to improve the comparability of various labeling schemes (Upham et al. 2011). The second issue with the labelling practice is the poor communication with consumers. A number of studies have identified that consumers are confused about the labeling information, even though they are willing to pay certain premium for carbon labeled product or service (Koistinen et al. 2013; Osman and Thornton 2019). In such context, research turns to improving visualization to help consumers better understand the utility of labeling scheme. For example, traffic light colored system was proposed to indicate intensity of product’ carbon footprint by using normalization method (Hornibrook et al. 2015; Thøgersen and Nielsen 2016). Whether such form of label suggested is effective to enhance communication still needs further validation. Besides, it is also worth noting that consumers may be irrational regarding environmental concerns (Lombardi et al. 2017). Conventional research methods, including questionnaires, focus group, in-depth interviews etc., have been widely employed to explore consumers perception and willingness to pay for carbon labelled product (Matukin et al. 2016). However, they may be limited by capturing responses based on consciousness (Van Gaal et al. 2011; Khushaba et al. 2013). Neuroscience is insightful to identify the conscious and subconscious responses thus to discriminate

16

1 Carbon Labeling and Its Related Issues

social consciousness and actual behaviors (Zhao 2019; De-Magistris et al. 2017). Such a tool has potential to investigate consumer’s perception towards different forms of labeling presentation. The third issue is the overlapped labeling policy. Taking food as an example, there were a number of labels presented on its package, such as information regarding organic, food miles, animal welfare, and carbon footprint (Galli et al. 2012). Various labels not only add complexity in the packaging design, but also give rise to the issues with respect to information credibility and reliability, even result in more confusions when consumer purchase product. Thus, the integration of various labeling policies is essential to ensure information coverage and improve labeling form of presentation (Shuai et al. 2014; Onozaka et al. 2016). The future research can be the optimization of life cycle assessment for labeling accreditation, improvement of labeling visualization for better expression, and normalization of various environmental label to promote sustainable consumption. How to implement carbon label standards as well as to homogenize carbon labeling schemes among different Countries should be investigated to provide insightful policy implications on sustainable consumption and production. Moreover, future studies may observe whether the carbon labeling scheme can trigger consumers to take pro-environmental purchasing actions. In such case, the research may deeply analyse the linkages between consumer and enterprises as well as between enterprises and government to implement carbon labelling scheme.

References Aertsens J, Verbeke W, Mondelaers K, Van Huylenbroeck G (2009) Personal determinants of organic food consumption: a review. Brit Food J 111:1140–1167 Aldy J, Pizer W, Tavoni M, Reis LA, Akimoto K, Blanford G, Carraro C, Clarke LE, Edmonds J, Iyer GC, McJeon HC, Richels R, Rose S, Sano F (2016) Economic tools to promote transparency and comparability in the Paris Agreement. Nat Clim Change 6:1000–1004 Brenton P, Edwards-Jones G, Jensen MF (2009) Carbon labelling and low-income country exports: a review of the development issues. Dev Policy Rev 27:243–267 Camilleri AR, Larrick RP, Hossain S, Patino-Echeverri D (2019) Consumers underestimate the emissions associated with food but are aided by labels. Nat Clim Change 9:53–58 Carbon Trust (2008a) Product carbon footprinting: the new business opportunity. Experiences from leading companies. The Carbon Trust, London Carbon Trust (2008b) Code of good practice for product greenhouse gas emissions and reduction claims (CTC745). The Carbon Trust, London Chen CM (2006) CiteSpace II: detecting and visualizing emerging trends and transient patterns in scientific literature. J Am Soc Inf Sci Tec 57:359–377 Chen CM (2017) Science mapping: a systematic review of the literature. J Data Inf Sci 2:1–40 Chen CM, Song M (2019) Visualizing a field of research: a methodology of systematic scientometric reviews. Plos One 14 Chen N, Zhang ZH, Huang SM, Zheng L (2018) Chinese consumer responses to carbon labelling: evidence from experimental auctions. J Environ Plan Man 61:2319–2337 Choi Y (2014) Global e-business management: theory and practice. Bomyoung-books Publishing Co, Seoul

References

17

Choi Y (2015) Introduction to the special issue on “Sustainable e-governance in Northeast-Asia: challenges or sustainable innovation’. Technol Forecas Soc Change 96: 1–3 Cohen MA, Vandenbergh MP (2012) The potential role of carbon labelling in a green economy. Energy Econ 34:S53–S63 DEFRA (Department for Environment, Food and Rural Affairs) (2009) Guidance on how to measure and report your greenhouse gas emissions. http://www.defra.gov.uk. Accessed 31 Oct 2011 De-Magistris T, Gracia A, Barreiro-Hurle J (2017) Do consumers care about European food labels? An empirical evaluation using best-worst method. Brit Food J 119:2698–2711 Deutz P, Neighbour G, McGuire M (2010) Integrating sustainable waste management into product design: sustainability as a functional requirement. Sus Dev 18:229–239 Falkner R (2016) The Paris Agreement and the new logic of international climate politics. Int Affairs 92:1107–1125 Fang K, Heijungs R, de Snoo GR (2014) Theoretical exploration for the combination of the ecological, energy, carbon, and water footprints: overview of a footprint family. Ecol Indic 36:508–518 Fujimori S, Kubota I, Dai H, Takahashi K, Hasegawa T, Liu J, Hijioka Y, Masui T, Takimi M (2016) Will international emissions trading help achieve the objectives of the Paris Agreement? Environ Res Lett 11 Galli A, Wiedmann T, Ercin E, Knoblauch D, Ewing B, Giljum S (2012) Integrating ecological, carbon and water footprint into a “footprint family” of indicators: definition and role in tracking human pressure on the planet. Ecol Indic 16:100–112 Gleim MR, Smith JS, Andrews D, Cronin JJ (2013) Against the green: a multi-method examination of the barriers to green consumption. J Retail 89:44–61 Gomi K, Shimada K, Matsuoka Y (2010) A low-carbon scenario creation method for a local-scale economy and its application in Kyoto city. Energ Policy 38:4783–4796 Grunert KG, Hieke S, Wills J (2014) Sustainability labels on food products: consumer motivation, understanding and use. Food Policy 44:177–189 Guenther M, Saunders CM, Tait PR (2012) Carbon labelling and consumer attitudes. Carbon Manag 3:445–455 Gupta S, Ogden DT (2009) To buy or not to buy? A social dilemma perspective on green buying. J Consum Market 26:376–391 Hornibrook S, May C, Fearne A (2015) Sustainable development and the consumer: exploring the role of carbon labelling in retail supply chains. Bus Strategy Environ 24:266–276 Hu AH, Chen CH, Lan YC, Hong MY, Kuo CH (2019) Carbon-labelling implementation in Taiwan by combining strength-weakness-opportunity-threat and analytic network processes. Environ Eng Sci 36:541–550 Iribarren D, Hospido A, Moreira MT, Feijoo G (2010) Carbon footprint of canned mussels from a business-to-consumer approach: a starting point for mussel processors and policy makers. Environ Sci Policy 13:509–521 Janssen M, Hamm U (2012) Product labelling in the market for organic food: consumer preferences and willingness-to-pay for different organic certification logos. Food Qual Prefer 25:9–22 Juvan E, Dolnicar S (2016) Measuring environmentally sustainable tourist behaviour. Ann Tourism Res 59:30–44 Kanada M, Fujita T, Fujii M, Ohnishi S (2013) The long-term impacts of air pollution control policy: historical links between municipal actions and industrial energy efficiency in Kawasaki City, Japan. J Clean Prod 58:92–101 Kane G (2010) The three secrets of green business: unlocking competitive advantage in a low carbon economy. Earthscan, London Khushaba RN, Wise C, Kodagoda S, Louviere J, Kahn BE, Townsend C (2013) Consumer neuroscience: assessing the brain response to marketing stimuli using electroencephalogram (EEG) and eye tracking. Expert Syst Appl 40:3803–3812

18

1 Carbon Labeling and Its Related Issues

Koistinen L, Pouta E, Heikkilä J, Forsman-Hugg S, Kotro J, Mäkelä J, Niva M (2013) The impact of fat content, production methods and carbon footprint information on consumer preferences for minced meat. Food Qual Prefer 29:126–136 Lazzarini GA, Visschers VHM, Siegrist M (2017) Our own country is best: factors influencing consumers’ sustainability perceptions of plant-based foods. Food Qual Prefer 60:165–177 Liu WL, Wang C, Xie X, Mol APJ, Chen JN (2012) Transition to a low-carbon city: lessons learned from Suzhou in China. Front Environ Sci Eng 6:373–386 Liu TT, Wang QW, Su B (2016) A review of carbon labelling: standards, implementation, and impact. Renew Sust Energ Rev 53:68–79 Lombardi GV, Berni R, Rocchi B (2017) Environmental friendly food. Choice experiment to assess consumer’s attitude toward “climate neutral” milk: the role of communication. J Clean Prod 142:257–262 Matukin M, Ohme R, Boshoff C (2016) Toward a better understanding of advertising stimuli processing exploring the link between consumers’ eye fixation and their subconscious responses. J Advertising Res 56:205–216 Messum D (2012) Creating change through carbon footprinting. http://www.carbontrust.com/news/ 2012/02/creating-change-through-carbon-footprinting Accessed 14 Mar 2017 Milovantseva N (2016) Are American households willing to pay a premium for greening consumption of information and communication technologies? J Clean Prod 127:282–288 Myeong S, Kwon Y, Seo H (2014) Sustainable E-governance: the relationship among trust, digital divide, and E-government. Sustainability 6: 6049–6069 Olive H, Volschenk J, Smit E (2011) Residential consumers in the Cape Peninsula’s willingness to pay for premium priced green electricity. Energy Policy 39: 544–550 Onozaka Y, McFadden DT (2011) Does local labelling complement or compete with other sustainable labels? A conjoint analysis of direct and joint values for fresh produce claim. Am J Agr Econ 93:693–706 Onozaka Y, Hu W, Thilmany DD (2016) Can eco-labels reduce carbon emissions? Market-wide analysis of carbon labelling and locally grown fresh apples. Agr Food Syst 31:122–138 Osman M, Thornton K (2019) Traffic light labelling of meals to promote sustainable consumption and healthy eating. Appetite 138:60–71 Peschel AO, Grebitus C, Steiner B, Veeman M (2016) How does consumer knowledge affect environmentally sustainable choices? Evidence from a cross-country latent class analysis of food labels. Appetite 106:78–91 Ramayah T, Lee JWC, Mohamad O (2010) Green product purchase intention: some insights from a developing country. Resour Conserv Recycl 54:1419–1427 Röös E, Tjarnemo H (2011) Challenges of carbon labelling of food products: a consumer research perspective. Brit Food J 113:982–996 Röös E, Sundberg C, Hansson P (2010) Uncertainties in the carbon footprint of food products: a case study on table potatoes. Int J Life Cycle Ass 15:478–488 Röös E, Sundberg C, Hansson P (2011) Uncertainties in the carbon footprint of refined wheat products: a case study on Swedish pasta. Int J Life Cycle Ass 16:338–350 Röös E, Sundberg C, Tidåker P, Strid I, Hansson PA (2013) Can carbon footprint serve as an indicator of the environmental impact of meat production? Ecol Indic 24:573–581 Rugani B, Vázquez-Rowe I, Benedetto G, Benetto E (2013) A comprehensive review of carbon footprint analysis as an extended environmental indicator in the wine sector. J Clean Prod 54:61– 77 Shi XP (2013) Spillover effects of carbon footprint labelling on less developed countries: the example of the East Asia summit region. Develop Policy Rev 31:239–254 Shi J, Visschers VHM, Bumann N, Siegrist M (2018) Consumers’ climate-impact estimations of different food products. J Clean Prod 172: 1646–1653 Shuai C, Ding L, Zhang Y, Guo Q, Shuai J (2014) How consumers are willing to pay for lowcarbon products? Results from a carbon-labelling scenario experiment in China. J Clean Prod 83:366–373

References

19

Song JS, Lee KM (2010) Development of a low-carbon product design system based on embedded GHG emissions. Resour Conserv Recy 54:547–556 Spaargaren G, van Koppen CSA, Janssen AM, Hendriksen A, Kolfschoten CJ (2013) Consumer responses to the carbon labelling of food: a real life experiment in a canteen practice. Sociol Ruralis 53:432–453 Tan CW (2009) Stimulating carbon efficient supply chains: carbon labels and voluntary public private partnerships Master diss. Massachusetts Institute of Technology, USA Tan MQB, Tan RBH, Khoo HH (2012) Prospects of carbon labelling—a life cycle point of view. J Clean Prod 72:76–88 Tan MQB, Tan RBH, Khoo HH (2014) Prospects of carbon labelling: a life cycle point of view. J Clean Prod 72:76–88 Thøgersen J, Nielsen KS (2016) A better carbon footprint label. J Clean Prod 125:86–94 Tian Y, Govindan K, Zhu Q (2014) A system dynamics model based on evolutionary game theory for green supply chain management diffusion among Chinese manufacturers. J Clean Prod 80:96–105 Upham P, Dendler L, Bleda M (2011) Carbon labelling of grocery products: public perceptions and potential emissions reductions. J Clean Prod 19:348–355 Van Gaal S, Lamme VA, Fahrenfort JJ, Ridderinkhof KR (2011) Dissociable brain mechanisms underlying the conscious and unconscious control of behavior. J Cogn Neurosci 23:91–105 Van Loo EJ, Caputo V, Nayga RM, Verbeke W (2014) Consumers’ valuation of sustainability labels on meat. Food Policy 49:137–150 Vermeir I, Verbeke W (2006) Sustainable food consumption: exploring the consumer “Attitude behavioral intention” gap. J Arg Environ Ethic 19:169–194 Wang CX, Wang LH, Liu XL, Du C, Ding D, Jia J, Yan Y, Wu G (2015) Carbon footprint of textile throughout its life cycle: a case study of Chinese cotton shirts. J Clean Prod 108:464–475 Wiedmann T, Minx J (eds) (2008) A definition of carbon footprint. Ecological economics research trends. New York, US Wu P, Feng Y (2012) Using lean practices to improve current carbon labelling schemes for construction materials—a general framework. J Green Build 7:173–191 Wu P, Xia B, Pienaar J, Zhao X (2014a) The past, present and future of carbon labelling for construction materials—a review. Build Environ 77:160–168 Wu P, Low SP, Xia B, Zuo J (2014b) Achieving transparency in carbon labelling for construction materials—lessons from current assessment standards and carbon labels. Environ Sci Policy 44:11–25 Wu P, Feng YB, Pienaar J, Xia B (2015) A review of benchmarking in carbon labelling schemes for building materials. J Clean Prod 109:108–117 Yang Y, Meng GF (2019) A bibliometric analysis of comparative research on the evolution of international and Chinese ecological footprint research hotspots and frontiers since 2000. Ecol Indic 102:650–665 Zander K, Hamm U (2010) Consumer preferences for additional ethical attributes of organic food. Food Qual Prefer 21:495–503 Zhao R (2019) Neuroscience as an insightful decision support tool for sustainable development. Iran J Public Health 48:1933–1934 Zhao R, Zhong S (2015) Carbon labelling influences on consumers’ behaviour: a system dynamics approach. Ecol Indic 51:98–106 Zhao R, Deutz P, Neighbour G, McGuire M (2012) Carbon emissions intensity ratio: an indicator for an improved carbon labelling scheme. Environ Res Lett 7:9 Zhao R, Zhou X, Han J, Liu C (2016) For the sustainable performance of the carbon reduction labelling policies under an evolutionary game simulation. Technol Forecast Soc Change 112:262– 274 Zhao R, Zhou X, Jin Q, Wang Y, Liu CL (2017) Enterprises’ compliance with government carbon reduction labelling policy using a system dynamics approach. J Clean Prod 163:303–319

20

1 Carbon Labeling and Its Related Issues

Zhao R, Han JJ, Zhong SZ, Huang Y (2018a) Interaction between enterprises and consumers in a market of carbon-labelled products: a game theoretical analysis. Environ Sci Pollut Res 25:1394–1404 Zhao R, Geng Y, Liu Y, Tao X, Xue B (2018b) Consumers’ perception, purchase intention, and willingness to pay for carbon-labelled products: a case study of Chengdu in China. J Clean Prod 171:1664–1671 Zhu J, Hua W (2017) Visualizing the knowledge domain of sustainable development research between 1987 and 2015: a bibliometric analysis. Scientometrics 110:893–914

Chapter 2

Consumer Behavior Towards Carbon Labeling Scheme

Abstract Consumers are open market drivers, and their acceptance of carbon labelled products is important in influencing the carbon labelling practice. This chapter aims to investigate consumers’ perceptions, purchase intentions and willingness to pay for carbon-labeled products. Although China has not implemented a carbon labelling scheme, it is critical to study the potential impact of such an initiative. This chapter takes Chengdu as a case city to discriminate consumers’ decisionmaking in buying carbon-labeled products by an empirical study combined with system dynamics (SD) simulation. It is expected that the results may provide insight into development of more efficient business strategies for sustainable consumption and production, to increase the potential for the carbon emissions reduction, thus to foster a carbon-labeling system in China. Keywords Carbon-labeled product · Purchase intention · Willingness to pay · System dynamics

2.1 Introduction Prior studies asked whether consumers might be indifferent to carbon-labeled products, as they were limited in providing individual utility (Spaargaren et al. 2013; Liu et al. 2016). For example, Borin et al. (2011) identified consumers’ primary preference for product quality and value, rather than labeling with environmental information. Due to unpredictable decision-making, consumers were regarded as less important roles in response to carbon-labeling schemes (Guenther et al. 2012). It’s necessary to investigate consumers’ behavior related to carbon-labeled products, to understand possible changes in terms of purchase intention and WTP, thus identifying the factors influencing the transition to a low-carbon society. As the final buyer of carbon-labeled products, consumers’ acceptance directly affects companies’ willingness to attach carbon labels to their products (Rijnsoever et al. 2015). Accordingly, an increasing number of studies have focused on consumers’ perception to the carbon labelling scheme or carbon labelled products. Related studies have shown that factors such as consumers’ awareness on carbon labels may give rise to various levels of perceived risk among consumers, which © Springer Nature Singapore Pte Ltd. 2021 R. Zhao and Y. Geng, Carbon Labeling Practice, https://doi.org/10.1007/978-981-16-2583-1_2

21

22

2 Consumer Behavior Towards Carbon Labeling Scheme

further affect their purchase intention and willingness to pay (Chekima et al. 2016; Maniatis 2016). Vanclay et al. (2011) found that sales for products with low-carbon labels would increase if an appropriate price mechanism is adopted. Hartikainen et al. (2014) further confirmed such a finding by conducting an online questionnaire survey to examine consumers’ perceptions on carbon-labeled products among Finnish consumers. Shuai et al. (2014a) performed a logistic regression analysis on consumers’ willingness to pay for carbon-labeled products, through which the level of education and monthly income were identified as the important influencing factors. Further, Zhao and Zhong (2015) found that the product’s premium, public’s level of education, and perceived consumer effectiveness were the main factors to influence whether consumers would buy carbon-labeled products. By using online choice experiments, Peschel et al. (2016) verified that a higher level of consumers’ knowledge would cause environmentally-friendly behaviors, in such a way that the consumers chose to buy carbon-labeled food. However, a number of studies considered that the current carbon labelling scheme did not provide a sufficiently meaningful message to the consumers, by which they may respond in a questioned way. It needs further improvement. For instance, Upham et al. (2011) found that consumers had difficulty in understanding the carbon emissions information that the labels conveyed. Through a questionnaire survey, Gadema and Oglethorpe (2011) found that most consumers felt confused when trying to comprehend carbon labels. Gössling and Buckley (2016) further found that consumers had a poor understanding on the use of carbon labels in the tourism industry. Sharp and Wheeler (2013) indicated that householders were expected the carbon labelling scheme to facilitate comparison of products by means of a traffic-light system. However, further study needed to validate whether the improved carbon labelling was easily understood by consumers. Prior studies on consumer behavior are mainly empirical, and based upon survey methods, including in-person interviews and online or offline questionnaires etc (Steiner et al. 2016). Kim and Trail (2010) took internal motivators, internal constraints, external motivators, and behavioural measures into account, to investigate the encouraging and discouraging factors that affected sports consumers’ behaviour. Dholakia et al. (2010) proposed a multi-dimensional approach from a consumer-centric view, to investigate consumer behaviour in a multichannel and multimedia retailing environment. Further, Upham et al. (2011) applied a focus group to examine the public’s attitudes toward carbon labeling in grocery products, and discussed the potential factors in influencing their purchasing motivations. Guenther et al. (2012)’s similar study also employed focus group to distinguish the consumers’ attitudes toward carbon-labeled products between United Kingdom and Japan. Echeverría et al. (2014) designed an open questionnaire to investigate Chilean consumers’ WTP for carbon-labeled products. Shuai et al. (2014b) and Vecchio and Annunziata (2015) further conducted experimental studies on consumers. While the former used a questionnaire survey to identify the influencing factors involved in low-carbon product purchasing, the latter focused on an auction experiment to assess price premiums for carbonlabeled food products. Li et al. (2017) built a logistic regression model based

2.1 Introduction

23

upon questionnaire survey to identify the main factors that influence consumers’ willingness to buy low-carbon products. Zhao et al. (2018a) extended the logistic regression analysis to understand how price premiums affect consumers’ WTP for carbon-labeled food products. In general, these studies indicated a need of improving carbon labels and uncovered that a strong market demand for carbon-labeled products had not been established. Also, these studies indicated that consumers have to afford product premium because producers have to bear extra costs due to accounting their carbon footprints (Li et al. 2016). However, few studies have been conducted to specify the level of the product’s premium, especially in a developing country, such as China, where normal consumers’ environmental awareness is still low. Under such a circumstance, this chapter aims to investigate consumers’ perceptions, purchase intentions and willingness to pay for carbon-labeled products. Although China has not implemented a carbon labelling scheme, it is critical to study the potential impact of such an initiative. Chengdu, a mega city locating in the southwestern China, it is a promising city in the inland China with rapid development. Given this background, this Chapter takes Chengdu as a case city to discriminate consumers’ decisionmaking in buying carbon-labeled products, thus providing theoretical and empirical support to develop low carbon consumption. In Sect. 2.2, a simplified flow of milk lifecycle is provided to show its embedded carbon emissions, including four stages, i.e., raw milk production, dairy product processing, product delivery and packaging waste disposal, to underpin the carbon labeling information that is demonstrated to interviewees for further studies on their perceptions and willingness to pay. Taking carbon-labeled milk as an example, Sect. 2.2 adopts questionnaire to investigate consumers’ perceptions, purchase intentions and willingness to pay. Section 2.3 designs a hybrid of an auction experiment and a consumption experiment to explore a homogenoeous consumers’ group, i.e. university students’ purchase intention and willingness to pay for a carbon-labeled food product. Section 2.4 provides a system dynamics approach to understand consumers’ perception to carbon labelled products indicated by the changes in the number of consumers. It is expected that these studies may provide insight into development of more efficient business strategies for sustainable consumption and production, to increase the potential for the carbon emissions reduction. Besides, the results may provide insightful evidence to foster a carbon-labeling system in China.

2.2 Carbon Footprint Assessment for Milk Product Food is a daily necessity, and its consumption and production have a remarkable influence on global carbon emissions (Virtanen et al. 2011; Wheeler and Von Braun 2013). China is the third world largest consumer of dairy products (Hagemann et al. 2012; Huang et al. 2014). The GHG emissions associated with dairy products are increasing annually, due to the continuous rise in consumer demand (Baek et al. 2014; Adler et al. 2015). With “green consumerism” gaining increasing influence on the market, development of low-carbon food is a practical need for the sustainable

24

2 Consumer Behavior Towards Carbon Labeling Scheme

development of food industry to reduce the GHG emissions, as well as to pursue longterm commercial success (Beske et al. 2014; Biggs et al. 2015). Carbon footprint, is an effective indicator to embody the low-carbon concept, regarded as the total carbon emission of a certain product or service during its entire life cycle (Vergé et al. 2013; Dong et al. 2014). In this section, the Tetra Pak 1L pure liquid milk is selected to assess its life cycle based carbon footprint. A number of studies have conducted measuring the carbon footprint of dairy products, by using the life cycle assessment (LCA). However, the conventional LCA, due to its complexities in required data acquisition, system boundary division etc., is difficult for engineers to implement in real applications (Chen and Corson 2014). This study is expected to provide a simplified assessment approach based upon a LCA framework of milk, mainly focuses on presenting its carbon emissions information to consumers, to help local dairy enterprises identify the most carbon-intensive sector of the whole life cycle, especially encourage the dairy enterprises with a higher environmental morality to have a product carbon labelling attempt, thus to provide effective measures for emissions reduction in dairy supply chain. LCA is to assess possible environmental impact based upon the quantitative survey of a product during its whole life cycle, by identifying environmental emissions of all materials and energy, to seek opportunities on improvement of product environmental performances (Huysveld et al. 2015). As defined by International Organization for Standardization (ISO), a precise LCA generally follows by four phases: Goal and scope definition, Life cycle inventory analysis (LCI), Life cycle impact assessment (LCIA), and Interpretation (Azarijafari et al. 2016). With regard to the conventional LCA, the process of life cycle impact assessment mainly uses different category indicators to elaborate results of the life cycle inventory (Nigri et al. 2014). However, only the product carbon footprint is considered in the impact category as the global warming potential, represented by kg CO2 e per kg emission. Other impacts, such as eutrophication, acid rain potential, toxicity etc., have been omitted in this study. As a lifecycle study may not always need to use impact assessment, the results of the LCI provide information of a product system, including all inputs and outputs in the form of elementary flows (Seppälä 2003), which is used to quantify the impact of carbon emissions in this study.

2.2.1 Definition of System Boundaries System boundary is a key component of LCA, which directly affects the assessment precision (Park et al. 2016). This study only focuses on the processes that directly contribute to pure milk production, that is, only the effects of energy and material input are taken into account, as shown in Fig. 2.1. The system boundary of the pure milk life cycle is thus simplified and divided into four stages, namely, raw milk production, dairy product processing, product transportation and packaging waste disposal. The carbon footprint is comprised of two parts, namely, direct GHGs emissions and indirect GHGs emissions, as shown in Eq. (2.1) (IPCC 2006):

2.2 Carbon Footprint Assessment for Milk Product

25

Fig. 2.1 System boundary of a pure milk product

G H G total = G H G dir ect + G H G indir ect

(2.1)

The direct emissions can be obtained by chemometrics, mass balance, or similar methods, and are calculated by using the following equation (IPCC 2006): G H G dir ect =

n 

Di × GW Pi

(2.2)

i=1

where i refers to the ith emissions source of the milk life cycle, D the activity level, GWP the global warming potential. The indirect emissions are calculated by using the following equation (IPCC 2006): G H G indir ect =

n 

Ai × E i

(2.3)

i=1

where i refers to the ith emissions source of the milk life cycle, A the activity level, which involves the amount of all resource and energy during the product life cycle (material input and output, energy use, transportation distance, etc.) E is the GHGs emission factor, which refers to the GHGs produced per unit activity level, derived from life cycle databases and industrial reports. The milk source base is located at Hongya Country, Southwestern Sichuan Province, China, about 147 kms far from Chengdu City, the provincial capital. The diary processing plant is located at Pixian, suburb of Chengdu City, which is about 175 kms far from the milk source base. The branded milk is mainly distributed to the central Chengdu City, about 40 kms distances from the diary processing plant.

26

2 Consumer Behavior Towards Carbon Labeling Scheme

Fig. 2.2 The geographic distribution of the local branded milk supply chain

The milk packaging waste is transported to the municipal landfill for final disposal, about 30 kms away from the central city. The detailed geographic distribution of the milk supply chain network is shown in Fig. 2.2.

2.2.2 Source of Inventory Data Table 2.1 shows the inventory data during the raw milk production stage, which are obtained by analogy to Hospido et al. (2003) on the milk LCA. The location of the raw milk production is dairy farms, where the main consumptions are fodder, electricity, and diesel. From a conversion of the results obtained by Hospido et al. (2003) on GHGs emissions factors from farm fodder and equipment disinfectant, a corresponding CO2 emission coefficient is obtained for the calculation. The electricity emission factor is based on the 2014 Baseline Emission Factors for Regional Power Grids in China, released by National Development and Reform Commission, of which the Central China power-grid emission factor is used in this study. The operating margin emissions factor is 0.972 t CO2 /MWh, and the build margin emissions factor is 0.47 t CO2 /MWh (NDRC 2014). Through the conversion, the power grid emission factor is 0.723 kg/kW·h. Table 2.2 shows the inventory data of the dairy processing stage. Specifically, a large amount of water is needed during the cooling and pre-heating of raw milk, as

2.2 Carbon Footprint Assessment for Milk Product

27

Table 2.1 Inventory data of the raw milk production Emissions type

Emissions source

Activity level

CO2 emissions factor

Source of emissions factor

Carbon footprint (g)

Energy

Electricity

0.047 kW·h

0.723 kg/kW·h

NDRC 2014

34

Diesel

3.68 ml

2.73 g/ml

IPCC 2006

10

Fodder

1290 g

0.403 g/g

Hospido et al. (2003)

522

Disinfectant

1.59 ml

1.79 g/ml

Hospido et al. (2003)

2.85

Water

2.66 L

0.009 g/L

Field investigation

0.25

Material

Table 2.2 Inventory data of the dairy processing Emissions type

Emissions source

Activity level

CO2 emissions factor

Source of emissions factor

Carbon footprint (g)

Energy

Diesel1

7.07 g

2.73 g/ml

IPCC 2006

22.8

Electricity

0.047 kW·h

0.723 kg/kW·h

NDRC 2014

33.5

Material

Cardboard paper

16.8 g

1.04 g/g

DEFRA 2012

17.4

Membrane

0.183 g

2.85 g/g

WRI 2004

0.522

Equipment cleaning

2.91 g

0.649 g/g

Field investigation

1.89

Tetra Pak

1.01 U

0.952 g/U

Field investigation

Water

4.41 L

0.094 g/L

Field investigation

96.1 0.415

well as the pre-heating and sterilization of the liquid milk. Electricity is mainly used for operating the equipment, and fuels are used to generate the steam required for the pre-heating and heating of the sterilizer (Riera et al. 2013). Cardboard paper is used to produce the outer packaging for the milk. The carbon footprint of product transportation is calculated by the transportation loads (tonne • km) multiplying the carbon emissions factor (Cai et al. 2012). According to the field investigation, light-weight gasoline truck in 2 tonnes of loading capacity is employed to transport the raw milk from the pasture to the processing plant, then to the distribution centre, and their distances are 175 kms and 40 kms, respectively. Heavy-weight diesel truck in 10 tonnes of loading capacity is employed to transport the packaging waste to the municipal landfill, with a distance of 30 kms. The CO2 emissions factors of the gasoline and diesel are measured by IPCC (2006), as 164 g per tonne km for 2 tonnes light-weight gasoline truck, 84.8 g per tonne km for 10 tonnes Heavy-weight disel truck, respectively, as shown in Table 2.3.

28

2 Consumer Behavior Towards Carbon Labeling Scheme

Table 2.3 Inventory data of the product transportation Transportation sub-stages

Activity level (t·km)

CO2 emissions factor(g/t·km)

Source of emissions factor

Carbon footprint (g)

Raw milk transportation

0.175

164

IPCC 2006;

28.7

Milk distribution

0.04

164

IPCC 2006;

Packaging waste transportation

0.03

84.8

IPCC 2006;

Cai et al. (2012) 6.55

Cai et al. (2012) 2.54

Cai et al. (2012)

Table 2.4 Inventory data of the packaging waste disposal Emissions type

Emissions source

Activity level

CO2 emissions factor

Source of emissions factor

Carbon footprint (g)

Recycling

Raw coal

9.03 g

2.69 g/g

IPCC 2006

24.3

2.09 kg/m3

IPCC 2006

9.68 × 10–10

Energy

Natural gas 4.63 × 10–10 m3

Landfill disposal

Crude oil

0.00776 g

3.07 g/g

IPCC 2006

0.0238

Electricity

0.0114 kW·h

723 g/kW·h

NDRC 2014

8.24

CO2

15.5 g

1.00 g/g

DEFRA, 2012

15.5

CH4

0.0843 g

25 g/g

DEFRA, 2012

2.11

12 g

1.31 g/g

Cherubini et al. (2009)

15.7

Recycling is selected for the packaging waste pre-treatment. Because of the separation technique for aluminum-plastic composites usually with a lower heating value, these containers are not suitable for incineration (Xie et al. 2011). Therefore, landfill disposal is the ultimate choice (Meneses et al. 2012). Data on the carbon emissions during the landfill stage are by analogy to Cherubini’s inventory of sanitary landfill disposal (Cherubini et al. 2009), shown in Table 2.4.

2.2.3 Results of Carbon Footprint Assessment The carbon footprints related to the four stages are listed in Table 2.5. The results show that for a typical 1L Tetra Pak of pure milk, the carbon footprint for its whole life cycle is 1120 g, of which 843 g is generated during the raw milk production, accounting for 75.27% of the total carbon footprint. The second stage is that of dairy processing, for which the carbon footprint is 173 g, accounting for 15.45%. The third stage of

2.2 Carbon Footprint Assessment for Milk Product Table 2.5 Carbon footprint of different lifecycle stages

29

Emission types

Carbon footprint (g CO2 per litre)

Percentage (%)

Raw milk production

843

75.27

Dairy processing

173

15.45

Product transportation

38

3.39

Packaging disposal

66

5.89

Total

1120

100

product transportation contributes 38 g of the carbon footprint, accounting for 3.39%. The carbon footprint in the stage of packaging waste disposal is 66 g, accounting for 5.89%. The results indicate that the largest contribution to the carbon footprint occurred at the acquisition stage of the raw milk, which accounts for 93.90% of the carbon footprint of its entire life cycle. The results are consistent with those of the study by González-García et al. (2013), which has indicated that raw milk production generated the highest carbon footprint (80–90%). Raw milk is considered taking from conventional dairy farming in the study, through which the carbon footprint at this stage accounts for 75.27% of the total. In the study by Hospido et al. (2003), the carbon footprint of the subsystem related to breeding during the farming stage accounts for 80.32% of the total carbon footprint, which is similar to the results of our study. Fodder has an impact on the GHGs emissions related to dairy cows, and especially the ratio of the ingredients in the mixed animal fodder has a significant influence (Castanheira et al. 2010). Based upon our field investigation, dairy cows are fed large amounts of coarse fodder due to its low cost. However, digesting this type of fodder may increase methane emissions (Muñoz et al. 2000; Hatew et al. 2016). Adjustment of the ratio of corn and coarse fodder in the animal feed may contribute to limiting the amounts of GHGs being emitted (Van-Middelaar et al. 2013). In addition, Excessive nitrogen fertilizer has been applied in the agricultural sector of China for a long time (Ha et al. 2015). The fertilizer in the soil would release N2 O by denitrification and in this way increase GHGs emissions (Rowlings et al. 2013). Thus, to create an effective balance between nutritional value and environmental impact is significant to be considered in fodder ingredients (Dutreuil et al. 2014). The carbon footprint of the raw milk production is 843 g, which is identified as the major source in the life cycle of milk. Specifically, farm fodder, such as corn and silage are the largest contributors (522 g), accounting for 46.61% of the total carbon footprint. The methane emissions of dairy cows are the second highest, with a carbon footprint of 273 g, accounting for 24.38% of the total carbon footprint, as shown in Fig. 2.3. This may be attributable to the ruminant digestive system of dairy cows, which may give rise to a large amount of methane (Wang et al. 2016).

30

2 Consumer Behavior Towards Carbon Labeling Scheme

Fig. 2.3 Carbon footprint of the raw milk production

In the dairy processing stage, the carbon footprint is 173 g, accounting for 15.45% of the total. Figure 2.4 shows that the major emissions source is the Tetra Pak production, which has a carbon footprint of 96 g, accounting for 8.58% of the total. Electricity and diesel energy consumption contribute 34 g and 23 g of carbon footprint, respectively, which account for 2.99 and 2.04% of the total. The carbon footprint of cardboard production is 18 g, accounting for 1.56% of the total. The carbon footprint contributions of water and membranes are relatively low (both < 0.1%). The product transportation is consisted by three sub-stages, as raw milk transportation, milk distribution and packaging waste transportation. Raw milk transportation has a carbon footprint of 29 g, accounting for 75.91% of the carbon footprint at this stage. Milk distribution and transportation of packaging disposal contribute 7 g and Fig. 2.4 Carbon footprint of the dairy processing

2.2 Carbon Footprint Assessment for Milk Product

31

3 g of carbon footprint respectively, which account for 17.35 and 6.74% of the carbon footprint at this stage, as shown in Fig. 2.5. At the packaging waste disposal stage, the carbon footprint is 66 g, accounting for 5.89% of the total. Figure 2.6 indicates that the major emissions source is raw coal consumption, which has a carbon footprint of 24 g, accounting for 36.87% of the carbon footprint at this stage. Landfill disposal is the second highest, which has a carbon footprint of 16 g, accounting for 23.82% of carbon emission at this stage. The carbon footprint associated with direct CO2 and CH4 emissions are 15 g and 2 g, which account for 23.52 and 3.2% of carbon emissions at this stage, respectively. Electricity consumption contributes 8.24 g of carbon footprint, which is 12.50% of the carbon footprint at this stage. As the consumption of crude oil and natural gases is relatively low, the contributions to the carbon footprint account for less Fig. 2.5 Carbon footprint of the product transportation

Fig. 2.6 Carbon footprint of the packaging waste disposal

32

2 Consumer Behavior Towards Carbon Labeling Scheme

than 0.05%. In China, landfilling is mostly used to dispose of paper-aluminumplastic composite packaging. However, it may cause the generation of harmful nondegradable substances (Woon and Lo 2013). To increase the recycling ratio may significantly mitigate the adverse environmental impact of the packaging waste disposal, in which development of aluminum-plastic separation technology is the effective way to promote the recycling of the waste Tetra Pak packaging.

2.2.4 Summary A simplified life cycle based assessment is employed to calculate the carbon footprint of a pure milk product, which is based upon the inventory analysis. The result indicates that the carbon footprint is mainly related to the production of raw milk at the farm, contributing 75.27% to the emissions, whereas dairy processing, product transportation and disposing of the packaging waste contribute 15.45%, 3.39% and 5.89%, respectively. For the simplified approach, the specific uncertainties related to the assessment results are: ➀ different division of system boundaries may lead to deviation in carbon footprint assessment. The system boundary of the study has been strictly defined, which only contains four procedures related to the pure milk product, i.e., raw milk production, dairy processing, product transportation and packaging waste disposal. However, the upper stream of the raw milk production, e.g., the raw milk source, as well as the downstream of product transportation, e.g., the product usage, has been deliberately omitted from the system boundary. ➁ the simplified approach mainly focuses on the inventory analysis for the impact assessment, which is quantified by the activity level multiplying the emissions factor. With regard to the activity level, it is closely related to the data acquaintance. However, there may be difficulties in obtaining the required data, thus limits the precision of calculated values. Emissions factor is another critical input for the impact assessment. Although some of the emissions factors have been measured by the field investigation, a number of factors, such as energy sources (electricity, diesels), fodder, disinfectant etc., are derived from the similar studies. There may be biases in the assessment results.This study may give insight to provide a transparent carbon emissions information to consumers, to encourage the dairy enterprises to implement emissions reduction related activities. The following section will take carbon-labeled milk products as an example to investigate consumer’s perception and willingness to pay.

2.3 A Questionnaire Survey Low carbon development has become a feasible strategy for the whole society to respond climate change. Carbon labeling is a useful tool to promote low carbon

2.3 A Questionnaire Survey

33

development. However, such an initiative has not been implemented in China. Therefore, it is crucial to examine its performance on consumers so that they can choose those low carbon products appropriately. In order to identify the key barriers of impeding the implementation of carbon labeling, this section aims to explore consumers’ perception of carbon-labeled goods, their purchase intentions, and willingness to pay for such products by conducting a questionnaire-based survey in the city of Chengdu, one of the most populated mega-cities in southwest China. Also, the unintended consequences by the application of a carbon labelling scheme are predicted, such as a product premium, so that appropriate policy recommendations can be proposed to help enterprises take positive actions toward implementing a carbon reduction labeling scheme.

2.3.1 Design and Collection of Questionnaire The questionnaire includes three parts with a total of 27 questions. The details of this questionnaire are provided in the Appendix. In particular, consumers’ perceptions and purchase intentions on carbon labeled products are gauged by questions with 3 multiple choices. Also, five questions are related with demographics so that more specific population information can be obtained. In addition, consumers’ consumption habits are measured by using different scales, including 19 questions. A 6-point Likert scale is selected to reduce statistic deviation and the scale is ranged from 1 to 6 (Chomeya 2010), including “completely disagree”, “strongly disagree”, “slightly disagree”, “slightly agree”, “strongly agree” and “completely agree”. Questions 1–4 examine consumers’ awareness on low-carbon behaviors in their daily lives, such as conserving water and saving electricity, waste management, and others (Bai and Liu 2013). Questions 5–9 measure the levels of consumers’ acceptance on carbon labeling (Weber et al. 2002). Questions 10–14 reflect the consumers’ effort levels on improving environmental performance (Webb et al. 2008, Zhao and Zhong 2015) and can measure perceived consumer effectiveness. Questions 15–19 assess consumers’ satisfaction on the carbon labeled product (Lee 2009) and can measure perceived benefits. Since a carbon labelling scheme has not been implemented in China, this study provides a brief introduction of carbon label with a representative figure proposed by the UK Carbon Trust (within 3 min), including its common definition and significance, its application to a specific product, e.g., milk in this investigation. A simplified flow of milk lifecycle that may contribute to a large proportion of carbon emissions is provided (as shown in Sect. 2.2), including four stages, i.e., raw milk production, dairy product processing, product delivery and packaging waste disposal, to help the interviewees better understand how the carbon information is generated in the label. Figure 2.7a shows the location of Chengdu city in China. In order to cover Chengdu’s major residential zones and ensure that the obtained questionnaire samples are well represented, Tianfu Plaza (one of the most famous shopping areas in Chengdu and also the crossing point of metro lines 1 and 2 where many residents transfer)

34

2 Consumer Behavior Towards Carbon Labeling Scheme

Fig. 2.7 The administrative region of Chengdu

is selected as the questionnaires distribution place, shown in Fig. 2.7b. Metro line 1 runs from the south to the north, while metro line 2 runs from the east to the west. The selection of such a crossing point is that the interviewees may come from different parts of Chengdu and thus can better representation the attitudes of local residents. In total, 2500 questionnaires were distributed with 1681 collected questionnaires, indicating a response rate of 67.2%.

2.3 A Questionnaire Survey

35

With the screening process, those incomplete questionnaires were removed from this study and finally 1132 valid questionnaires were remained. However, the homogeneity issue has to be considered. For instance, the total samples show that most interviewees are below the age of 35. Especially, students account for a large proportion of the interviewees, about 51.6%. A possible reason may be that several universities, middle and high schools locate near this site, e.g., Sichuan University, Southwest Jiao Tong University, Chengdu Traditional Medical University, etc. This subway station serves as the major transfer station for the students. However, these interviewees with similar backgrounds may have similar understandings on some items, which may lead to uncertainty of the results, or even bias. This issue needs to be solved by diversifying the selection of interviewees. But due to higher population mobility in the investigation site, it is difficult to diversity the interviewees. In order to reduce the impact of interviewees’ homogeneity, all the valid questionnaires have been screened again. First, the Excel random number generator was employed to ensure stochastic sampling, by which each interviewee was assigned with a random number, and then ranked in a sequence from the highest to the lowest. Then, in order to better select the final samples, we decided to follow Chengdu’s demographic features so that perspectives from different stakeholders with different ages, income, gender and other backgrounds can evenly be investigated. For instance, the minimum samples for the interviewees who received education from senior high school was set as 25.4% of the total finally selected samples since residents with the same education level in Chengdu accounted for 25.4% of the total population. Such a measure helps reflect more comprehensive opinions on low carbon consumption from more stakeholders, rather than only focusing on some special groups. Particularly, some elders may not often take subways and therefore missed our interviews. But their percentage in the overall population is quite large and their views on such a matter should be included in our analysis. In addition, the number of finally investigated samples should be as 5–10 times of the number of questions so that valid statistical analysis can be conducted (Dolnicar et al. 2016). By taking such considerations, 453 of 1132 questionnaires were finally selected for data analysis. The software of SPSS 19.0 was used to perform reliability and validity tests. Reliability test assesses whether a questionnaire is stable and reliable, in which the Cronbach’α coefficient is adopted to evaluate the measurement scale’s reliability (Inrig et al. 2012). The Cronbach’α reliability coefficient of this study is 0.867, higher than the acceptable threshold of 0.7, indicating that the scale of 19 questions gauging individuals’ consumption habits are reliable. On the other hand, validity test mainly measures a scale’s construct validity through exploratory factor analysis (Ruscio and Roche 2012). A prerequisite to perform factor analysis is to satisfy Kaiser–Meyer–Olkin (KMO) measure of sampling adequacy analysis and the Bartlett’s sphericity test (Izquierdo et al. 2014). The KMO value of the 19 items in this study is 0.867, higher than the standard threshold of 0.7. The level of significance of the Bartlett’s sphericity test is 0, indicating that the questionnaire’s items are appropriate for conducting factor analysis. The dependent variables are set as the consumers’ perceptions, purchase intentions, and willingness to pay for carbon-labeled products. To clarify their influences,

36

2 Consumer Behavior Towards Carbon Labeling Scheme

an exploratory factor analysis is performed on the 19 items. The principal components are extracted and combined with the demographic variables so that the independent ones can be incorporated into the logistic regression model. Finally, regression analysis is conducted on the factors influencing the dependent variables. Factor analysis is a statistical technique to extract common factors from a group of variables (Frazier et al. 2013). The items measuring consumption habits in this study are defined as x 1 , x 2 ,…, x 19 , respectively. The factor model can be expressed as: xi = ai1 f 1 + ai2 f 2 + · · · + aim f m + u i

(2.4)

where f 1 , f 2 ,…,f m are the common factors of the measured items; ui is the special factor (that is, the residual term); and α ij is the factor loading. Through the factor analysis, the information of the observed variables can be transformed into the factor values f 1 , f 2 ,…, f m . Subsequently, these factors can be adopted to replace the original observed variables in order to perform the logistic regression analysis. When the model’s dependent variables are non-continuous binary or multivariate, traditional regression analysis is not applicable. Logistic regression adopts the idea E(y) becomes a linear function of x 1 , x 2 ,…, x k . of logit transformation, so that ln 1−E(y) Thus, it becomes possible to process binary dependent (or multivariate) variables (Ozdemir and Altural 2013). The equation of logistic regression simulation is given as follows:  p = β0 + βi f i 1− p i=1 m

ln

(2.5)

where f i is the value of factor, m is the number of factors, and β is regression coefficient to be estimated; p is the probability when the dependent variable y equals to 1.

2.3.2 Descriptive Statistic Results Table 2.6 presents the descriptive statistical results regarding consumers’ perceptions, purchase intentions, and willingness to pay for carbon-labeled products. Regarding awareness on carbon-labeled products, 78.1% of respondents have never heard of carbon labels. This confirms the fact that carbon-labeled products have not yet been promoted in China. In terms of purchase intentions, taking milk products as an example, 70.9% of respondents would consider buying carbon-labeled milk, and 83.7% of them even can accept a level of premium below 0.6 RMB. These findings indicate that although these respondents have an overall low awareness of carbon

2.3 A Questionnaire Survey

37

Table 2.6 General descriptive statistics Question item

Sample

Percent(%)

perception

Have you ever heard of No carbon-labeled products? Yes

354

78.1

99

21.9

Purchase intention

Would you buy carbon-labeled milk products?

No

132

29.1

Yes

321

70.9

What is an acceptable price of carbon-labeled milk products (given that the original price of milk is 3.00 RMB/box)?

Unwilling to pay more

Willingness to pay

82

18.1

Extra 0.03–0.3 RMB

164

36.2

Extra 0.33–0.6 RMB

133

29.4

Extra 0.63–0.9 RMB

38

8.4

Extra 0.93–1.2 RMB

36

7.9

labeled products, most of them would purchase them even with a bit higher price if such products occur in the markets. Factor analysis was conducted on the 19 question items. Results show that the cumulative variance was 56.64% after four principal components were extracted, where questions X1-4, X5-9, X10-14, and X15-19 corresponded to the factors from f1 to f4 respectively, shown in Table 2.7.

2.3.3 Regression Results The casual relationship of the logistic regression analysis is divided into three stages. The first stage mainly focuses on consumers’ perception on carbon-labeled products, set as a binary dependent variable. Four significant factors, representing consumers’ low-carbon awareness (X1 to X4), uncertainly of acceptance (X5 to X9), perceived effectiveness (X10 to X14), perceived benefits (X15 to X19), together with the demographic variables, are deemed as the independent variables of the logistic regression model, given in a bullet list: • • • • •

Low-carbon awareness. Uncertainty of acceptance. Perceived consumer effectiveness. Perceived benefits. Demographic variables.

Table 2.8 shows that only uncertainty of acceptance has a significant impact on the perception of carbon-labeled products. Other factors, such as demographics, do not significantly influence the perception. This finding is consistent with the results from Weber et al. (2002), who carried out an assessment on perceptional risks by using five psychological scales (financial decision-making, health and safety, leisure, ethics, and social decision-making). As a result, consumers’ uncertainty of acceptance may

38

2 Consumer Behavior Towards Carbon Labeling Scheme

Table 2.7 Component matrix of factor analysis on the 19 question items Component f1 X5

0.824

X6

0.789

X7

0.772

X8

0.723

X9

0.629

f2

X15

0.740

X16

0.710

X17

0.700

X18

0.654

X19

0.615

f3

X10

0.774

X11

0.731

X12

0.699

X13

0.593

X14

0.560

f4

X1

0.698

X2

0.672

X3

0.671

X4

0.415

Table 2.8 Logistic regression analysis on consumers’ perception of carbon-labeled products Coefficient Subjective risk

Standard error Wald

0.720∗∗∗ 0.149

df Sig

23.243 1

Exp(B)

0.000 2.055

Perceived benefits

0.047

0.127

0.134 1

0.714 1.048

Perceived consumer effectiveness

0.014

0.122

0.013 1

0.908 1.014

Low-carbon awareness

− 0.099

0.114

0.750 1

0.387 0.906

Gender

− 0.074

0.251

0.088 1

0.767 0.928

Age

− 0.063

0.145

0.188 1

0.664 0.939

0.025

0.180

0.019 1

0.889 1.025 0.638 1.042

Level of education Occupation

0.042

0.088

0.222 1

Income

− 0.150

0.104

2.066 1

0.151 0.861

Constant

− 0.967

0.735

1.731 1

0.188 0.380

Note Chi-square = 31.134, df = 9, sig = 0.000, Nagelkerke R2 = 0.102. The classified predicted value was 78.6% *Significant the p < 0.1; ** Significant at p = 0.05, ***Significant at p = 0.01.

2.3 A Questionnaire Survey

39

have significant impact on raising perception of carbon-labeled products. Indirectly, it also shows that those who are more willing to try new things and brands in their daily lives are more likely to be aware of carbon-labeled products. The second phase of regression analysis focuses on exploring the major factors influencing consumers’ willingness to pay for carbon-labeled products. In other words, consumers’ intention to buy carbon-labeled products is set as the binary dependent variable. The four significant factors, demographic variables, and the degree of perception on carbon-labeled products are set as the independent variables, given as follows: • Low-carbon awareness. • Uncertainty of acceptance. • Perceived benefits. • Perceived consumer effectiveness. • Demographic variables. • Degree of perception on carbon labeled products. From the regression coefficients and levels of significance shown in Table 2.9, it is clear that five factors, including level of education, age, low-carbon awareness, perceived benefits, and perceived consumer effectiveness, significantly affect consumers’ intentions to buy carbon-labeled products. Perceived benefits and lowcarbon awareness are the most crucial factors. These findings are similar to the conclusions from Maniatis (2016), in which he concluded that four factors determined consumers’ decisions on buying green products, including perceived benefits for the environment, economic benefits, green reliability, and the product’s green appearance. Consumers’ education levels, ages, and perceived consumer effectiveness are Table 2.9 Logistic regression analysis on consumers’ intention to buy carbon-labeled products Coefficient

Standard error

Subjective risk

− 0.012

0.122

Perceived benefits

0.626∗∗∗

0.115

Perceived consumer effectiveness

0.193∗

0.113

Low-carbon awareness

0.402∗∗∗

Perception of carbon-labeled products

Wald

df

Sig

Exp(B)

0.010

1

0.919

0.988

29.427

1

0.000

1.871

2.940

1

0.086

1.213

0.111

13.169

1

0.000

1.494

0.209

0.282

0.548

1

0.459

1.233

Gender

0.158

0.238

0.442

1

0.506

1.172

Age

− 0.229∗

0.137

2.796

1

0.094

0.795

Level of education

0.417∗∗

0.166

6.324

1

0.012

1.518

Occupation

− 0.002

0.084

0.001

1

0.978

0.998

Income

0.001

0.098

0.000

1

0.993

1.001

Note: Chi-square = 58.840, df = 10, sig = 0.000, Nagelkerke R2 = 0.174; the classified predicted value was 70.9% *Significant the p < 0.1; ** Significant at p = 0.05, ***Significant at p = 0.01.

40

2 Consumer Behavior Towards Carbon Labeling Scheme

the next most important factors. The higher the education level and the younger the consumer, the more willing the consumer is to buy carbon-labeled products. Similar findings can be found from Shuai et al. (2014a). The third phase involves the selection of consumers who intends to purchase carbon-labeled products. In order to understand the level of premium that consumers are willing to pay for carbon-labeled milk products, four significant factors, demographic factors, and the degree of perception of carbon-labeled products are further employed as the independent variables, given as follows: • Low-carbon awareness. • Uncertainty of acceptance. • Perceived benefits. • Perceived consumer effectiveness. • Demographic variables. • Degree of perception on carbon labeled products. Since willingness to pay can be represented by the degree of premium, 5 levels of premium are defined. The level of premium at five different levels is then taken as the dependent variable. As the parallel lines testing models fail to run when all the variables are input, the model is run again after eliminating all insignificant variables. The P value of parallel lines testing is 0.09, larger than 0.05, meaning that all regression equations are mutually parallel. In other words, the regression coefficient of the independent variable is not related to the cut-off point, thus satisfying the pre-requisite of conducting ordinal logistic regression (Tehrani and Ahrens 2016). Table 2.10 presents the results of consumers’ willingness to pay for carbon-labeled products, where the perceived consumer effectiveness has the most significant influence on willingness to pay, followed by the uncertainty of acceptance. The corresponding regression coefficient is positive, which further explains that consumers who believe that purchasing carbon-labeled products can improve existing environmental conditions are more willing to spend money on such products. With regard to the demographic variables, occupation and income level have significant influence on one’s willingness to pay for carbon-labeled products. In terms of occupation, students have the strongest impact on the results, while freelancers have the weakest effect. In terms of income level, only the group with a monthly income of less than 1500 RMB reaches statistical significance. This might be due to the fact that milk is a commodity with weaker need elasticity. Consumers (especially those with higher incomes) are generally not very sensitive to price fluctuations (Mostafa 2016).

2.3.4 Summary Carbon labeling is a quantitative expression on one product’s carbon footprint and can inform the consumers about one product’s carbon information, which may encourage them to select low-carbon products (Edwards-Jones et al. 2009). With a carbon labeling system, sources of carbon emissions can be more transparent and induce

2.3 A Questionnaire Survey

41

Table 2.10 Ordinal regression analysis of consumers’ willingness to pay for carbon-labeled products Coefficient Standard error Wald df Significance Subjective risk

0.129∗∗

0.065

3.892 1

0.049

Perceived benefits

0.026

0.067

0.146 1

0.702

Perceived consumer effectiveness

0.177∗∗∗

0.069

6.615 1

0.01

Low-carbon awareness

− 0.037

0.070

0.277 1

0.599

Occupation 1: Student

0.973

1.129

0.742 1

0.389

Occupation 2: Freelancer

0.579

1.146

0.256 1

0.613

Occupation 3: Teacher, doctor, scientific 0.767 staff (etc.)

1.164

0.434 1

0.510

Occupation 4: Civil servant or public officer

0.918

1.175

0.610 1

0.435

Occupation 5: Business employee

0.622

1.152

0.291 1

0.589

Occupation 6: Retired

0a

–

Income 1: < 1500 RMB

− 0.783∗∗∗ 0.295

7.040 1

0.008

Income 2: 1500–3000 RMB

− 0.213

0.226

0.889 1

0.346

Income 3: 3000–4500 RMB

− 0.287

0.229

1.573 1

0.210

Income 4: 4500–6000 RMB

0.175

0.254

0.476 1

0.490

Income 5: > 6000 RMB

0a

–

–

–

–

0

0

–

Note Joint function: Auxiliary logarithm-logarithm *Significant at p < 0.1; ** Significant at p = 0.05, ***Significant at p = 0.01 a The parameter is redundant, set at 0

the changes of public consumption behaviors (Wu et al. 2014). Such a measure can help the promotion of low carbon products (Li and Colombier 2009). However, it has not been practiced in China (Liu et al. 2016). That is why the consumers in Chengdu generally have lower perceptions of carbon-labeled products according to this investigation. The results also indicate that the consumers’ perceived benefit significantly influences purchase intentions, followed by the education level and low carbon awareness. Concerning willingness to pay, perceived consumer effectiveness, occupation, and level of income are the significant influencing factors. Such findings indicate that the consumers with better education and positive environmental awareness may be the major group on buying the carbon labeled products. Thus, it may be crucial to first influence such consumers so that they can interact with other consumers by sharing their purchasing experiences.

42

2 Consumer Behavior Towards Carbon Labeling Scheme

2.4 A Purchase Decision-Making Experiment People’s purchase intention is affected by various factors, such as external stimuli (e.g., labeling information and price), socio-demographic variables (e.g., gender and age), and psychological factors (Hernandez-Ortega, 2011; Trope and Liberman 2010). To understand how consumers may respond to a carbon-labeling scheme, the product selection criteria calls for most consumers to be familiar with the products’ attributes and price (Echeverría et al. 2014). Socio-demographic factors, including age, educational level, and environmental awareness, among others, have yielded substantial results regarding consumers’ purchase intentions and WTP for carbonlabeled products (Grunert et al. 2014; Shuai et al. ). University students (or students enrolled in a university) have a keen environmental consciousness. They are receptive to new trends and inclined to be involved in sustainable consumption (Emanuel and Adams 2011). They constitute a big group of future low-carbon consumers and contribute to the diffusion of environmental-friendly consumption (Shuai et al. 2014b). This section applies the principle of homogeneity to the experimental design, that people prefer interacting with those who are similar to themselves (Triandis 1989). The University Students in this study is taken as a homogeneous group, to investigate their purchase intention and willingness to pay for carbon-labeled products. This study thus takes this demographic as a potential consumer group to observe whether they have similar preferences, thus to provide insight into low-carbon consumption. Given that university students are typically familiar with milk attributes, such as nutritional components and taste (Zhao and Zhong 2015), this study exclusively focuses on (fluid/liquid) cow’s milk (abbreviated to “milk” hereafter). It first roughly identifies the attributes potentially influencing consumers’ purchase intention through focus group discussions. An auction experiment is then conducted to determine the influence of the attributes identified in focus group discussions. A field consumption experiment is further carried out to verify the auction experiment results and determine students’ WTP in terms of an acceptable range of price premiums. Existing carbon label-related studies heavily use structured or semi-structured interviews and questionnaire surveys (favored data collection methods in social sciences) to obtain necessary information from respondents (detailed in the ensuing section). Such methods are reasonably useful but are by no means above reproach. They are still subject to many shortcomings (e.g., high time and monetary costs, interview bias, time-lapse considerations, and potential deviation from actual observed human behaviors), some of which may considerably decrease the validity and reliability of the research findings (Khushaba et al. 2013). As such, this study takes a motivational experiment design and conducts behavioral experiments, including an auction experiment and a consumption experiment, to examine the purchase intention and WTP for carbon-labeled food products of students in a Chinese university. This study, which can sufficiently capture actual observed human behaviors, is expected to enrich our understanding of university students’ carbon label-related behaviors and serve as a valuable supplement to the existing literature. Additionally,

2.4 A Purchase Decision-Making Experiment

43

this study is conducive to developing a carbon-labeled product market, encouraging more residents to participate in low-carbon consumption, and promoting food sustainability. Locally, it provides policy insights into developing a carbon-labeling system in China.

2.4.1 Experimental Design This study selects fluid milk among food products as an example to identify the possible influence attributes of consumers’ purchase intentions through a focus group interview. Consumers find milk as a daily consumed product to be affordable, and they are familiar with its attributes, such as nutritional components and taste (Zhao and Zhong 2015). Such influence attributes as attribute variables are then applied to the auction experiment to determine core attributes of influence based on the auction results. A field consumption experiment further verifies the auction results and determines consumers’ WTP in terms of an acceptable range of product premiums. Figure 2.8 illustrates the roadmap designed for this study. The experimental group includes 282 students from a key university in Chengdu City, China, who selected the “Environmental Protection and Sustainable Development” module in the academic years of 2015 and 2016. These students were interested in sustainable development, and had expressed a relatively strong environmental consciousness. To a certain extent, they served as pioneers of low-carbon consumption, in order to interact with the consumers and enhance their environmental consciousness. The module was simultaneously offered in four classes (sequenced by curricular numbers), and the experimental class was selected using single-attribute complete randomization, as Table 2.11 illustrates. From the class with the smallest random number (curricular number: B3709), 24 students were selected to form 4 focus groups, and 6 students in each focus group were interviewed in a semistructure, which is based upon a pre-determined set of open questions (indicated in Table 2.12) with the opportunity for the interviewer to extract necessary information in a short period of time (Bryman 2017). The moderator controlled the field procedure during the one-hour seminar, and asked the respondents the questions related to carbon-labeled products for their discussion. Subsequently, the discussion results

Fig. 2.8 Experiment design

44

2 Consumer Behavior Towards Carbon Labeling Scheme

Table 2.11 Sampling results of the experimental classes Curricular number

Class size

Random number

B2131

67

0.606

B2132

88

0.896

B2133

60

0.976

B3709

67

0.062

Table 2.12 Question design for the focus group Number

Question

1

Have you heard of carbon labeling?

2

What do you think of utility of carbon labeling?

3

Would you buy milk for your daily consumption?

4

Why would you choose to purchase fluid milk?

5

What concerns you when purchasing milk?

6

Would you choose to purchase carbon-labeled milk?

7

Why would you choose to purchase carbon-labeled milk?

8

Are you concerned with labeling information when purchasing milk?

9

Do you think carbon-labeled milk will be more expensive?

10

Would you buy carbon-labeled milk if its price were higher than non-carbon-labeled milk?

11

If so, to what extent you would accept an increased price?

were applied as the attribute variables to the milk auction, and the students involved in the largest random-numbered class were set as the bidders. Students in the class with the largest random number (curricular number: B2133) were chosen as bidders. The entry of the experiment was fully voluntary. Anyone willing to participate was included in the experimental group. Before conducting the experiment, essential requirements were informed, including research aim, duration, procedure, and ethics. The willingness to participate in environmental sustainability research is observed to be rooted in the students’ mindsets. Finally, fifty participants were recruited, as the other ten students exited the auction experiment for personal reasons. The participants were randomly assigned into three groups (numbered as 1, 2, and 3, respectively). The auction experimenttook into account four crucial variables identified in focus group discussions: packaging, taste, nutrient ingredients, and carbon-labelling information. The auctioned products contained both carbon-labeled and non-carbonlabeled 250 mL box-packed pure milk, 243 mL box-packed chocolate milk, 250 mL box-packed high-calcium and low-fat milk, and 240 mL bagged pure milk. The labels’ carbon emissions information was derived from the results by Zhao et al. (2012) and Zhao et al. (2018b).

2.4 A Purchase Decision-Making Experiment

45

As formerly noted, fifty students acted as the bidders (purchasers), while the course lecturer served as the auctioneer. A bidding price range was predetermined, in which milk products’ market prices were used as the price mean and five price levels were set above and below the mean, respectively. Therefore, each type of milk product experienced ten rounds of bidding. In each round of auction, the price increases by 0.1 Chinese Yuan (CNY). Every time the auctioneer quoted a price, the bidders were given seven seconds to decide whether to bid. If they were willing to bid, they could raise their hands. To mitigate the possible interference of social reactions on individual decision making, all participants were asked to respond simultaneously. In such a case, each respondent does not have enough time to interact with others, so the interactive influence can be properly controlled. If a bidder did not participate in a round of bidding, he or she could participate in other rounds. When the bidding prices reached the predetermined ceiling price, the auction was declared as complete. Multi-round auctions may result in the affiliation effect. In other words, in multiround auctions, if some bidders discover others’ overbidding, they will follow others to push prices up. Consequently, the final price may deviate from the products’ actual value. Therefore, this study only employed a one-round auction. This study conducted a field consumption experimentto verify the auction’s experimental results and explore university students’ consumption behaviors influenced by the price differentials between the carbon-labeled and the non-carbon-labeled milk. A small shop on the campus was rented to sell three types of common milk products with the same brand (huorun), which differed in price, packaging, and taste. Both carbon-labeled and non-carbon-labeled products were sold. Six samples of milk products were selected, marked as A, a; B, b; and C, c, respectively, as Fig. 2.9 demonstrates. As China has no carbon-labeling scheme, a self-designed carbon label was marked on the milk packaging, as presented in Fig. 2.10. Carbon label was first proposed by the UK Carbon Trust in 2006 and presented in a form of numerical value, which is based on the public emission reduction pledge of a specific product or service according to lifecycle based carbon emissions assessment (Zhao et al. 2012; Liu et al. 2016). Our self-designed carbon label is noted as a “CO2 ” shape; specifically, the “O” is replaced by a leaf, which signifies low carbon and environmental friendliness, nature, and health. The data at the upper-right corner of the carbon label indicates the carbon emissions across the product’s lifecycle. This study estimated its emissions as approximately 200 g for carbon-labeled products in Milk

A

a

products Carbon labeling

Fig. 2.9 Experimental milk products

B

b

C

c

46

2 Consumer Behavior Towards Carbon Labeling Scheme

Fig. 2.10 Prototype of the self-designed carbon label

Table 2.13 Price premiums in different sales periods

Period

Price premium (CNY per Product)

1

0

2

0.1

3

0.2

terms of previous findings, due to the time and cost restrictions on the experimental milk’s carbon emissions throughout its lifecycle (Corrigan and Rousu 2006). The sale’s target group involved the 282 students in the 4 experimental classes who did not participate the focus group discussions and the auction experiment. Sixteen students were unwilling to join, and the other 192 were finally engaged in the selling experiment. Each student in the experiment received one cash coupon with his or her student ID, and this coupon allowed them to purchase one specific type of milk product once. After a student purchased a milk product, his or her cash coupon would be collected, and the purchased milk product type would be recorded. The selling experiment was implemented in three consecutive weeks, and one week represents one period. This study designed three price premium levels, as Table 2.13 notes, based on prior studies of the price premiums of carbon-labeled milk (Zhao and Zhong 2015).

2.4.2 Experiment Results Data collected in the auction experiment were first analyzed using descriptive statistics and followed by an independent-sample t-test in the SPSS 19.0 software to investigate whether and how the carbon label impacted the bidders. The experimental data were further subject to a partial correlation analysis to identify the correlation between each control variable and the percentage of bidders. Figure 2.11 presents the variation and indicates the error (or uncertainty) in bidders’ percentage for the four types of milk products, carbon-labeled and non-carbon-labeled. For the same auction price, the percentage of bidders for the carbon-labeled milk is generally higher than that for the non-carbon-labeled milk. With the increase in prices, a decrease in the bidders’ percentage for both the carbon-labeled and the non-carbon-

2.4 A Purchase Decision-Making Experiment

47

Fig. 2.11 Variations in the percentage of bidders for the four milk products

labeled milk is seen, which is highly reasonable. Figure 2.11 also shows that for the same auction price, the difference in bidders’ percentages for the carbon-labeled and the non-carbon-labeled milk is statistically significant in most cases, although there are still some price levels where error bars overlap. Table 2.14 shows the maximum differences in the percentage of bidders and their corresponding auction prices. Table 2.15 lists the test results regarding the percentage of bidders for the carbonlabeled and non-carbon-labeled milk. Their variances are homogeneous, implying that they are capable of a mean comparison. The difference in the means is 0.1598, indicating that the mean percentage of the bidders for the carbon-labeled milk is Table 2.14 Correlation between different variables and bidder percentages

Variable

Partial correlation coefficient

Price

−0.876

Taste

0.867

Nutritional ingredients

0.684

Package

−0.592

Carbon label

0.513

Hypothesis: Variances are unequal

Hypothesis: Variances are equal

238

237.046

−4.306

−4.306

0.655

0.419*

0.000**

0.000**

−0.15983

−0.15983

Difference in mean

the 95% confidence level, the original hypothesis is supported and the variances are homogeneous **At the 95% confidence level, the original hypothesis is disproved and the means are significantly different

* At

Percentage of bidders

Sig (Bilateral)

df

t

F

Sig

T-test of variance equation

Levene’s test of variance equation

Table 2.15 Independent sample test

0.03712

0.03712

Standard error

−0.23296

−0.23295

Lower limit

−0.08671

−0.08671

Upper limit

Confidence level of difference

48 2 Consumer Behavior Towards Carbon Labeling Scheme

2.4 A Purchase Decision-Making Experiment

49

slightly higher than that for the non-carbon-labeled milk. This reveals that a carbonlabeling scheme had a certain positive influence on the university students’ bidding behaviors. This study presented the influence on the university students’ bidding behaviors— exerted by each attribute as reflected in the survey of focus groups to analyze the correlation between a specific attribute and the percentage of bidders. Partial correlation analysis was employed to measure the degree of association between the two random variables, i.e., price, taste, nutritional ingredients, package, and carbon label with the bidders’ percentage respectively, which is effective to remove the impact of remained controlling variables. Table 2.14 reveals that price strongly correlates with the percentage of bidders, indicating that price is the core influencing attribute of the consumers’ consumption behaviors, which is consistent with conclusions from prior studies (Chekima et al. 2016). In contrast, carbon labeled products slightly affects the percentage of bidders; evidently, consumers’ primary concern involves the products’ basic attributes (Zhao and Zhong 2015). Only when product attributes satisfy their consuming needs will they consider additional information, such as carbon labels (Hartikainen et al. 2014). The milk prices negatively correlate with the percentage of bidders, in that the rise in milk price may result in a decrease in bidders. A similar tendency is found in the milk packaging, which also indicates a negative correlation with bidders. Tam et al. (2016) confirmed this result, in that packaging convenience was important for university students while purchasing food: the more complicated the food packaging, the fewer students welcomed it. In contrast, carbon labeling, taste, and nutritional ingredients positively correlate with the percentage of bidders; specifically, the carbon-labeling scheme, diversity of milk tastes, and improved nutritional value increased the number of bidders. Table 2.16 displays the sales volume of the carbon-labeled and non-carbon-labeled milk in the three periods. No price difference occurred between the three types of milk in the first period; 84.6% of the effective participants chose to purchase the carbon-labeled milk, and the rest chose to purchase the non-carbon-labeled milk. The university students were more inclined to buy carbon-labeled milk. When the price premium was 0.1 yuan in the second period, 78.9% of the effective participants chose to purchase carbon-labeled milk, and the rest chose to purchase the non-carbonlabeled milk. When the price premium was 0.2 yuan in the third period, 56% of the effective participants chose to purchase the carbon-labeled milk, and the rest chose to purchase the non-carbon-labeled milk. The percentage of purchases of carbonlabeled milk did not significantly differ from that of the non-carbon-labeled milk, which indicated a significant decrease in the university students’ WTP due to the price premium. Additionally, the sales volume of the No. 1 or No. 2 milk was greater than that of the No. 3 milk in each period, indicating that the university students were inclined to purchase the box-packed milk products under the same price premium. Figure 2.12 indicates that the sales volume gradually decreased for both the carbon-labeled and non-carbon-labeled milk in each period, with the increased price premium. The carbon-labeled milk’s sales volume more noticeably decreased than that of the non-carbon-labeled milk. In contrast, the non-carbon-labeled milk’s sales

7

4

The 21 second week

The third week

5

8

30

The first week

No. 2

No. 3

9

28

38

6

22

28

4

5

4

10

27

32

3

17

19

3

4

2

6

21

21

A a Total B b Total C c Total (Carbon-labeled) (Non-carbon-labeled) (Carbon-labeled) (Non-carbon-labeled) (Carbon-labeled) (Non-carbon-labeled)

No. 1

Table 2.16 Sales volume of different types of milk in the field consumption experiment

50 2 Consumer Behavior Towards Carbon Labeling Scheme

2.4 A Purchase Decision-Making Experiment

Sales

51

Carbon labeled product

90 80 70 60 50 40 30 20 10 0 1st week

2nd week

3rd week

Period Fig. 2.12 Variations in sales volume for carbon-labeled and non-carbon-labeled milk

volume increased slightly in the second period, and decreased in the third period. The carbon-labeled milk’s sales volumes in the first and second periods were far greater than that of the non-carbon-labeled milk. Further, a significant decrease occurred in the third period regarding the sales volume of both the carbon-labeled and non-carbon-labeled milk, with the former slightly higher than the latter. The results indicate that once the milk in the experiment was labeled with carbon footprint information, the price premium of 0.1 yuan was generally acceptable, and the premium deviation was 3.2%. This phenomenon can be explained by Lombardi et al.’s conclusion, in that university students are deemed as rational consumers with no independent sources of income (Lombardi et al. 2017). As milk is a daily necessity, the students are acquaintance with its attributes, including the average retailing price and the possible price fluctuation. If the students perceive such price premium extremely surpasses their expected payment, they are unwilling to buy the carbon labeled milk even though they commonly have high educational levels and environmental awareness.

2.4.3 Summary This study sequentially implemented focus group discussions, an auction experiment, and a field consumption experiment to determine factors influencing the purchase of carbon-labeled products (more specifically, milk). The focus group discussions reveal that the purchase of the carbon-labeled milk is mainly influenced by the price, taste, packaging, and nutritional value. The auction experiment and the field consumption experiment further indicate that price is the primary factor influencing consumers’ intention to buy carbon-labeled milk. Consumers are unwilling to buy if they feel that the price premium is too high (Lin and Huang 2012; Moser 2015). Specifically,

52

2 Consumer Behavior Towards Carbon Labeling Scheme

university students with no independent financial capabilities are vulnerable to price premiums in choosing carbon-labeled products (Lam 2014; Echeverría et al. 2014). The consumption experiment indicates the carbon level’s acceptability to consumers (e.g., acceptable maximum price differential). When the price premium is 0.1 Yuan, many students still choose to purchase the carbon-labeled milk. This result implies that most consumers are insensitive to the price premium within a specific price range. Echeverría et al. (2014) drew a similar conclusion: Chilean consumers were willing to pay 29% more for milk and 10% more for bread than their respective average prices. Moreover, when the price premium is increased from 0.1 Yuan to 0.2 Yuan, an apparent decrease in the carbon-labeled milk sales is witnessed, which suggests that multiple premium points may cause a sudden change in purchase intention. Moreover, the carbon-labeled milk has a higher sales volume than the non-carbon-labeled one, within a premium range acceptable to the experimental group. This indicates that the carbon-labeled products are potentially in demand among the university student group. We have noticed that a number of the students are accompanied with their roommates or classmates to our rented shop during the period of consumption experiment, to consult with their companions regarding the purchase of an experimental milk product. Companions are a trusted group of consumers, so their opinions may influence individual consumers’ perceptions of products’ utilities, and even cause changes in the consumers’ preferences (Lindsey-Mullikin and Munger 2011; Henderson and Beck 2011). Additionally, the companions share their purchasing experiences with each other, and positive comments will increase the individuals’ confidence in their purchasing behaviors (Hart and Dale 2014). In contrast, the companions’ indifference to or negative comments about products will somewhat diminish the consumers’ purchase intentions (Borges et al. 2010). Moreover, carbonlabeled milk has a higher sales volume than the non-carbon-labeled milk, within the premium range acceptable to the experimental group. This indicates that the carbonlabeled products are potentially in demand among the university student consumer group. We have also found several students inquire the meaning of our proposed labelling scheme attached to the milk products during the consumption experiment, when they are faced to which experimental milk may be purchased. Similar studies revealed such dilemma as Upham et al. (2011) and Zhao et al. (2012) identified that current carbon reduction labelling scheme was difficult for consumers to imagine a given quantity of CO2 emission and its embodied environmental impact. In this case, transition of the current carbon labelling system to be more transparent may be a useful way to reinforce its environmental communication.

2.5 A System Dynamics Simulation Most of the existing studies are based upon an emprical analysis and cannot demonstrate the consumers’ intention in a quantitative way. This section provides a system dynamics approach to better understand consumers’ perception to carbon labelled

2.5 A System Dynamics Simulation

53

products, by visually showing how the number of consumers is varied, to investigate the key factors influencing on consumers’ purchasing choice (e.g. the environmental attitude, environmental consciousness, perceived consumer effectiveness etc.), thus to validate the empirical findings of prior studies. Two case scenarios are presented to demonstrate the application of the proposed approach, in which the consumers’ preferences to the carbon labelled and non-labelled milk, as well as the carbon labelled milk with different fat contents are simulated. According to the five-stage model in consumer buying process, consumers are classified into potential consumers, ordinary consumers and loyal consumers (Mowen and Minor 2001). It is assumed in the study that the ‘potential consumers’ may not have intention of purchasing the carbon labelled products at the initial stage, whilst the ‘loyal consumers’ purchase them repetitively (at least twice). In addition to the former two categories, others are deemed as the ‘ordinary consumers’. Basically, consumer is mainly affected by the external stimulus (e.g., price, promotion etc.) and individual characteristics (socio-demographic factors, environmental attitude, environmental awareness, environmental knowledge, perceived risk etc.) to arouse buying decisions (Kotler and Armstrong 2010). It is believed that price is a key factor to affect consumers’ preferences to the green products (Rezai et al. 2011). Due to additional cost for better raw materials, green products are usually sold at a higher price as compared with conventional products (Ling 2013). Furthermore, consumers are willing to buy the environmental labelled products with the price slightly rising, due to appeal of perceived better quality (D’souza et al. 2006; Barnard and Mitra 2010). This can be verified by Hassan and Nor (2013), who have confirmed that consumers may be charged even higher for green products, based on the survey of purchase intention of lead-free electronics. For any new product, which may sell at a price premium, advertisement is necessary for its promotion. For instance, the main information channels are TV and newspapers for consumers to purchase electronic products in Malaysia (Pillai and Meghrajani 2013). The socio-demographic factors (gender, education level etc.) are indicated to be correlated to the consumption of environmental-friendly products, especially the education background shows a close correlation (Bonti-Ankomah and Yiridoe 2006; Makower and Pike, 2009). For example, an investigation conducted in Portugal showed that the majority of the green activists were aged 25–34 or 45–54, and most of them were well-educated with high income (Do Paço et al. 2009). Meyer and Liebe (2010) further confirmed that the high income group would be willing to buy environmentally friendly products. Gender did not play a key role in the attitude towards green consumption (Chen and Chai 2010; Pillai and Meghrajani 2013). However, Noor et al. (2012) argued that female consumers could have a greater preference on the green products based on an investigation at hypermarkets in Malaysia, whilst household income had the least influence. Fisher et al. (2012) further suggested consumer behaviour should be described in specific scenarios, in order to exam the availability of demographics. The environmental attitude is defined as “a psychological tendency expressed by evaluating the natural environment with some degree of favour or disfavour”

54

2 Consumer Behavior Towards Carbon Labeling Scheme

(Milfont and Duckitt 2010). Milfont (2007) indicated that the environmental attitude is related to the perceived environmental threat, which may affect the environmental behaviour. For the environmental behaviour of individuals, the environmental attitude is the primary explanatory variable. Any change of the environmental attitude may have a reflection on the environmental behaviour (Steg and Vlek 2009). Chen and Chai (2010) divided the environmental attitude into three dimensions, i.e., environmental protection, governmental role and personal norm, in which the latter two had an impact on consumers’ attitude to green products. Based on a multiple linear regression analysis of the surveyed data, the personal norm, or the moral obligation in some perspectives, was identified as the major influencing factor, whilst the environmental protection was the least. However, Gadenne et al. (2011) demonstrated that consumers with positive environmental attitude would prefer environmentally friendly actions such as recycling, energy saving etc., and buy the green products at a highly affordable price. Similar studies were found by Birgelen et al. (2009) and Dono et al. (2010), who confirmed that the positive environmental attitude resulted in eco-friendly actions. The environmental consciousness, deemed as an awareness of environment, acts in a way for natural preservation and environmental protection, with minimum impact on earth (Mainieri et al. 1997). Arslan et al. (2012) applied structural equation model to investigating the environmentally conscious purchasing behaviour of university students. The model result showed that the environmentally conscious consumption was significantly affected by environmental attitude and awareness of green products. In another study, the influences of environmental consciousness, perceived product attributes and attitudes towards product on the purchase intentions were investigated (Assarut and Srisuphaolarn 2012). The results indicated that the individual environmental consciousness could indirectly affect purchase intentions in terms of perceived attributes of the green product. However, the perceived product attributes acted on the purchase intentions directly. It was also suggested that the promotion of green products should not only take the overall product attributes into account, but also establish effective communication with the potential consumers. Environmental knowledge contains a general knowledge, e.g., facts, concepts, and relationships of the natural environment and major ecosystems (Fryxell and Lo 2003). Peattie (2010) considered the environmental knowledge as a key factor to drive the green consumption, whilst Bartiaux (2008) argued that there was no cause-andeffect relationship between environmental knowledge and environmentally friendly action, or at least their relationships were uncertain (Zsóka 2008). Tan (2011) further indicated that the environmental knowledge, perceived environmental threat and consumer effectiveness could positively impact on the green purchasing behaviour, while the environmental attitude served as a moderating variable. The perceived risk mainly involves two components, namely a ‘chance’ focused on probability or uncertainty, and a ‘danger’, i.e., the negative consequences (Mitchell 1999; Campbell and Goodstein 2001). Therefore, any factor related to the uncertainty or the consequences may result in the perceived risk of consumers, including personal characteristics, experience, knowledge, trust etc. (Lobb et al. 2006). For most of the risk aversions, any slightly increase in the perceived risk may give rise to a negative

2.5 A System Dynamics Simulation

55

effect on their purchasing behaviour (Tonsor et al. 2009). In addition, the perceived risk plays a key role in price premium. For instance, a survey of the consumers’ preference to products with reused or recycled content indicated that the higher perceived functional risk led consumers to choose new product instead (Essoussi and Linton 2010). Proposed by Kinnear et al. (1974), the perceived consumer effectiveness (PCE) refers to a measure that an individual believes his or her effort to mitigate environmental impact. For example, Joonas (2008) indicated that the PCE could better understand the driver of green consumption, compared with consumers’ income. Webb et al. (2008) found that the PCE was highly related to the consumers’ social responsibility. For environmentally conscious consumers, who are aware of green consumption, may likely be influenced by the corporate social responsibility (CSR). As the strong social norms are encouraged by pro-environmental innovations, people tend to take more environmentally friendly actions (Ozaki 2011). From the above analysis, the study selects the following factors for the development of SD model, i.e., perceived risk, education level, advertisement, price, income, environmental consciousness, environmental attitude, environmental knowledge, perceived consumer effectiveness, social environmental responsibility, personal value, product carbon emissions, acceptance of carbon labelling.

2.5.1 Model Formulation System dynamics is an approach, combined quantitative and qualitative analysis together, in order to understand the transformation of a complex system (Sumari et al. 2013). Proposed by Prof. Forrester in 1956, the system dynamics has been widely used in various fields, e.g., public policy making (Ghaffarzadegan et al. 2011), supply chain management (Vlachos et al. 2007), environmental protection (Mukherjee et al. 2013; Manasakunkit and Chinda 2013). In this study, the application of system dynamics is applied to modelling consumers’ perception to the carbon labelled products. Two scenarios are built for the simulation of consumers’ purchasing behaviour in terms of modelling variation of the three defined categories of consumers. Scenario 1: model consumers’ responses (in terms of number variation) to the carbon labelled and non carbon labelled milk, where the price is different, shown in Fig. 2.13. Scenario 2: model consumers’ responses (in terms of number variation) to a range of similar products, as carbon labelled milk, i.e., milk with different fat contents and carbon emissions, whole milk, semi-skimmed and skimmed milk, shown in Fig. 2.14. In order to understand the transformation of consumers’ purchasing decision making, the variation in the number of the three defined categories of consumers is used as an indicator in the SD model. Let variation 1, variation 2, variation 3 and variation 4 denote the number of consumers change in a given time period, from potential to ordinary, ordinary to loyal, ordinary to potential, loyal to potential,

56

2 Consumer Behavior Towards Carbon Labeling Scheme

v ariation4

uncertainty

v ariation3

ordinary consumers

potential consumers

price f actor

perception risk

publicity f actor

loy al consumers

v ariation2

v ariation1

risk attitude income f actor

perceiv ed ef f ectiv eness

price premium income

degree of inv olv ement product price without carbon labelling

adv ertisment product Inf ormation

env ironmental education

gov ernmental education media broadcasting

consumers' interaction

altriustic env ironmental v alue

env ironmental responsibility

env ironmental knowledge

~ product satisf action rate

ecological v alue

~

~

~

product price with carbon labelling

carbon labelling f actor

v alue

env ironmental attitude

env ironmental awareness

univ ersity prof essional education

env ironmental organisation

public awareness

total consumers

self ish env ironmental v alue

acceptance of carbon labelling

Fig. 2.13 The SD model of the first scenario

v ariation4

uncertainty

v ariation3

ordinary consumers

potential consumers

perception risk

publicity f actor

price f actor

env ironmental awareness

risk attitude income f actor

perceiv ed ef f ectiv eness

income

degree of inv olv ement

product Inf ormation

product price without carbon labelling

product price with carbon labelling

env ironmental education ~

consumers' interaction

gov ernmental education media broadcasting

health risk

~

~

env ironmental organisation

f at content

Fig. 2.14 The SD model of the second scenario

self ish env ironmental v alue

univ ersity prof essional education

public awareness

total consumers

ecological v alue

altriustic env ironmental v alue

env ironmental responsibility

env ironmental knowledge

~ product satisf action rate

carbon labelling f actor

v alue

env ironmental attitude

price premium

adv ertisment

loy al consumers

v ariation2

v ariation1

acceptance of carbon labelling

carbon emission f actor

2.5 A System Dynamics Simulation

57

respectively. Their corresponding transformation mechanisms are shown in Table 2.17. In addition, the transformation indicators are driven by intermediate variables, such as environmental awareness, publicity factor, perception risk etc., as shown in Table 2.18. Table 2.17 Mechanism of consumers’ transformation Indicator

Transformation mechanism

Variation 1

IF ((environmental awareness + publicity factor-perception risk–price factor) < 0), THEN (0.01*number of potential consumers), ELSE ((environmental awareness + publicity factor–perception risk–price factor)/2* number of potential consumers)

Variation 2

IF ((environmental awareness + carbon labelling factor + consumer value + environmental attitude- income factor) < 0), THEN (0.01*ordinary consumers), ELSE ((environmental awareness + carbon labelling factor + consumer value + environmental attitude–income factor)/3* number of ordinary consumers)

Variation 3

Number of ordinary consumers–variation 2

Variation 4

Uncertainty*number of loyal consumers

Table 2.18 Measurement of the intermediate variables Intermediate variable

Measurement

Environmental awareness

MEAN (environmental attitude, consumers’ interaction, environmental knowledge)

Publicity factor

MEAN (advertisement, consumers’ interaction)

Price factor

MEAN (price premium, income factor)

Perception risk (Scenario 1)

IF ((degree of involvement–product information–risk attitude) < 0), THEN (0.01), ELSE (degree of involvement–product Information–risk attitude)/2

Perception risk (Scenario 2)

IF ((degree of involvement + health risk–product information–risk attitude) < 0.01), THEN (0.01), ELSE (degree of involvement + health risk–product Information–risk attitude)/2

Environmental attitude

MEAN (perceived effectiveness, environmental knowledge, environmental responsibility)

Consumer value

(Altruistic environmental value + ecological value-selfish environmental value)/2

Income factor

IF ((product price with carbon labelling/income) > = 0.6), THEN (1), ELSE (product price with carbon labelling/income)

Carbon labelling factor (Scenario 1)

Acceptance of carbon labelling

Carbon labelling factor (Scenario 2)

Acceptance of carbon labelling–carbon emissions factor

58

2 Consumer Behavior Towards Carbon Labeling Scheme

The intermediate variables are determined by decision variables, i.e., premium ratio (The price of carbon labelled product divided by the price of unlabelled product), income, attitude towards product risk, public awareness, education level and perceived consumer effectiveness. Their corresponding values are derived from the similar studies by analogy and field investigation, which are shown in Table 2.19. Here, the target product, as the 250 millilitre boxed milk, is investigated in a local supermarket at Chengdu, Sichuan province, China to model the possible consumers’ purchasing behaviours. In addition, the total number of consumers is set as 10,000, and it is assumed that 1 box of the milk is purchased per capita daily.

2.5.2 Simulation Results The variation in number regarding the three defined categories of consumers is shown in Fig. 2.15. The number of potential consumers decreases gradually, but the number of loyal consumers keeps increasing. For the ordinary consumers, the number increases and then gradually decreases in the early stage of the simulation. This can be explained as a considerable proportion of the potential consumers transforms into the ordinary ones in a short interval, thus to result in a rapid increase of the ordinary consumers. Once the transition from the ordinary to the loyal consumers is above that from the potential to the ordinary consumers, decrement of the ordinary consumers occurs. Ultimately, a dynamic equilibrium is achieved that all the three defined categories of consumers remain in the same level for a long term period. In order to figure out the possible influences on purchasing of the carbon labelled milk, a single factor analysis is implemented by incorporating the six decision variables, as shown in Table 2.20. As the loyal consumers in this scenario are indicated as the consumers with a larger preference to the carbon labelled milk, the following analysis mainly focuses on their number variation. Figure 2.16 reflects that the price influencing on the loyal consumers may be negligible if the premium rate is less than 1.1. However, the number of the loyal consumers decreases drastically, when the premium rate is 1.15. Thus, the price acted on consumers’ behaviour is mainly depending upon the critical premium (i.e., the highest price that consumers are willing to pay). When the price of the carbon labelled milk is lower than the critical premium, the price has little effect on the consumer behaviour, and vice versa. Figure 2.17 shows that income has less impact on the change of loyal consumers, which is in agreement with Noor et al. (2012), but contrary to Meyer and Liebe (2010). A possible reason is that milk is the daily necessity with a low price, which accounts for a less proportion of consumers’ expenses. Even for the group of people with lowincome (1500 CNY per month), the milk is still affordable for daily consumption according to the field investigation. Figure 2.18 shows that the risk attitude has no effect on the change in the loyal consumers, which is opposite from what Tonsor et al. (2009) have identified. A

2.5 A System Dynamics Simulation

59

Table 2.19 Measurement of the decision variables Decision variable

Measurement

Numerical values

Price of the carbon labelled and unlabelled milk

It is indicated that 71.6% of the consumers are willing to pay more for green products, in which 26.4% claims to buy the products even the price is higher, about 5% to 10% (Zhao et al., 2014). As there is no carbon labelled milk commercially available in China market, the price of is set as 10% higher than the original price of the investigated milk

3 Chinese Yuan (CNY) per litre in terms of the market survey in Chengdu city. 3.3 CNY per litre for the carbon labelled milk

Income

The income refers to the disposable income per capita of urban residents, which is comprised of the final consumption expenditure, other non-compulsory expenses and savings (National Bureau of Statistics of the People’s Republic of China 2013). The annual disposable income per capita for urban residents in China is 26,955 CNY (National Bureau of Statistics of the People’s Republic of China 2014)

2246 CNY monthly

Risk attitude

The risk attitude is the chosen 0.2 (range from 0 to 1.0; response of a person to the higher value indicates expected utility (Wärneryd, stronger willing to take risks) 1996). Previous study indicated that the consumer’s attitudes towards risks are 5.56 out of 7.0, where 7.0 means the absolute risk aversion (Gao 2009)

Public awareness

The public awareness of environmental issues. Yin (2010) identified that 63.5% of the surveyed consumers paid greater attention to the current environmental issues

0.6 (range from 0 to 1.0; higher value indicates higher awareness)

(continued)

60

2 Consumer Behavior Towards Carbon Labeling Scheme

Table 2.19 (continued) Decision variable

Measurement

Numerical values

Education level

The percentage of college-educated citizens in China (Census Office of the People’s Republic of China, 2010). According to the population census, it is calculated as 0.089

0.09 (range from 0 to 1.0)

Perceived consumer effectiveness

It is shown that the respondents with highly perceived consumer effectiveness only accounts for 26.68% of the investigated people (Qing et al., 2006)

0.3 (range from 0 to 1.0)

Fat content (Scenario 2)

The fat content in every 100 g milk

For whole milk 4%, semi-skimmed milk 2%, skimmed milk 0.1% (Zhao et al., 2012)

Fig. 2.15 Variation in number of the defined consumers in the first scenario Table 2.20 The values variation of the decision variables Decision variable

Value 1

Value 2

Value 3

Premium rate

1.05

1.1

1.15

Income (CNY per month)

1500

2750

4000

Risk attitude

0.2

0.5

0.8

Public awareness

0.3

0.6

0.9

Education level

0.1

0.4

0.7

Perceived consumer effectiveness

0.2

0.5

0.8

2.5 A System Dynamics Simulation

61

Fig. 2.16 Variation in number of the loyal consumers with different premium rates

Fig. 2.17 Variation in number of the loyal consumers with different income

possible reason is that milk is a daily consumed product, by which most of consumers are familiar with its basic attribute, e.g. nutrition, contents etc. Thus, the risk towards the carbon labelled milk does not highly influence on the consumers’ behaviour. Figure 2.19 illustrates that the number of loyal consumers increases as the public awareness rises, which is similar to what Assarut and Srisuphaolarn (2012) have identified.

62

2 Consumer Behavior Towards Carbon Labeling Scheme

Fig. 2.18 Variation in number of the loyal consumers with different risk attitudes

Fig. 2.19 Variation in number of the loyal consumers with different public awareness

Figure 2.20 shows that the education level is positively correlated to the variance in number of the loyal consumers, which is in agreement with Do Paço et al. (2009), that the green consumers are identified as well-educated, who may better understand the environmental knowledge and have a positive environmental attitude. Figure 2.21 reflects that the higher the perceived consumer effectiveness is, the larger the number of the loyal consumers is, which is in accordance with the results of Webb et al. (2008). The higher perceived consumer effectiveness may lead to a

2.5 A System Dynamics Simulation

63

Fig. 2.20 Variation in number of the loyal consumers with different education level

Fig. 2.21 Variation in number of the loyal consumers with different perceived consumer effectiveness

stronger belief that individuals trust they can significantly enhance the environmental protection. According to the previous definitions of the three defined consumers, it is further established in the scenario that the ‘potential consumers’ are intended to purchase the milk without carbon label, whilst the ‘loyal consumers’ only buy the carbon labelled milk. For the ‘ordinary consumers’, the probability of choosing either milk to buy is identical. The sales amount of the carbon labelled and carbon unla-

64

2 Consumer Behavior Towards Carbon Labeling Scheme

Fig. 2.22 Variation in sales volume of the carbon labelled and unlabelled milk in the first scenario

belled milk are simulated, shown in Fig. 2.22. As the intensity of promotion rises, the consumers will give priority to purchasing the carbon labelled milk, thus to increase the corresponding sales volume. In the second scenario, the consumer loyalty is further defined as the loyalty to the low-carbon products. It was proposed that the carbon emissions intensity ratio decreased as the fat content reduced, as far as a range of milk, i.e. whole milk, semi-skimmed and skimmed milk were concerned (Zhao et al. 2012). Thus, the loyal consumer is deemed as the consumer who purchases the skimmed milk with carbon labelling at least twice. The simulation result of the second scenario is shown in Fig. 2.23, where 1, 2 and 3 are used to represent the number of consumers who buy whole milk, semi-skimmed, skimmed milk, respectively. All the defined consumers are inclined to purchase the skimmed milk, and especially the number of loyal consumers keeps increasing, which reflects that the consumers may give preference to the low carbon milk. It is further indicated that the fat contents and the associated carbon emissions have a significant impact on the consumer’s decision making, with little variance of sales price. For the second scenario, the ‘potential consumers’ was given an equal probability in their decision making, i.e., 50%, either purchasing the whole or semi-skimmed milk. The ‘ordinary consumers’ purchase the three kinds of milk randomly. The ‘loyal consumers’ buy both the semi-skimmed and skimmed milk, but give priority to the latter one. The sales volume of the skimmed milk surpasses the whole and semi-skimmed milk, as shown in Fig. 2.24.

2.5 A System Dynamics Simulation

65

Fig. 2.23 Variation in number of the defined consumers with different fat contents of the carbon labelled milk in the second scenario

66

2 Consumer Behavior Towards Carbon Labeling Scheme

Fig. 2.24 Variation in sales volume of the whole, semi-skimmed and skimmed milk with carbon labelling in the second scenario

2.5.3 Summary The simulation results show that both of the two scenarios have the similar trends varied by the number of three defined consumers, and the number of loyal consumers accounts for the largest proportion, indicating that the consumers may give preference to the carbon labelled products. This can be verified by analogy to Chou (2013) in green consumption study, which shows that 66.3% of the investigated consumers are intended to buy the green products. The critical premium, public awareness, education level and the perceived consumer effectiveness have considerable influences on the purchasing behaviour, whilst the effects of income and attitude towards risk can be negligible. Hence, emphasis should not only centre on raising the consumers’ environmental awareness, but also develop the market of carbon labelled products by introducing possible target consumers. It is further suggested that the consumers with better education, higher perceived consumer effectiveness and positive environmental awareness, may be considered as the target consumers, who are encouraged to be involved in the trial of buying the carbon labelled products, thus to accelerate the low carbon consumption.

Appendix Questionnaire. (a) Question items of consumers’ perceptions on carbon-labeled products and demographics.

Appendix

67 Order Question item

Options

Consumers’ perceptions on 20 carbon-labeled products

Have you ever heard about carbon-labeled products?

Never heard of them Have heard of them

21

If there is a carbon-labeled milk product (this label promotes low carbon living), would you buy it?

No Yes

22

If you are willing buy carbon-labeled milk (you selected option B in the previous question), what is an acceptable extra amount that you are willing to pay? (given that each carton of milk costs 3.00 RMB)

Unwilling to pay more 3.03−3.30 RMB 3.33–3.60 RMB 3.63–3.90 RMB 3.93–4.20 RMB

23

Your gender:

Male Female

24

Your age:

18–25 B. 26–35 C.36–45 D. 46–55 E.56 or above

25

Your highest level of education (or currently attending):

A.Lower middle school or less B. High school, vocational middle school, and advanced vocational training C. College and undergraduate D. Postgraduate or above

26

Your occupation:

Student B. Freelancer or self-employed C. Teacher, doctor, scientific staff (etc.) D. Civil servants or public officer E. Business employee F.Retired

27

Your monthly income:

1500 RMB or less 1500–3000 RMB 3000–4500 RMB 4500–6000 RMB 6000 RMB or above

Demographics

(b) Question items measuring consumers’ consumption habits.

68

2 Consumer Behavior Towards Carbon Labeling Scheme

Number

Question item

Note

1

I have the habit of switching off the lights when leaving a room

Low-carbon awareness

2

I re-use tap water. For example, after washing my face, I re-use the same water to wash my feet, then again to flush the toilet

3

I keep plastic shopping bags and re-use them

4

I collect empty drink bottles and sell them to waste collectors

5

I enjoy accepting new ideas and items

6

I would rather spend more time making comparisons before shopping instead of regretting my decision afterwards

7

I want to understand the relevant information on carbon labels accurately before shopping

8

I am willing to try a new brand

9

Buying carbon-labeled milk products can satisfy my sense of novelty and curiosity

10

I am aware how my consumption behavior impacts the environment

11

I am willing to select only environmentally-friendly products from now on, even if it is inconvenient for me

12

I am willing to make personal sacrifices to enhance the quality of the environment, even if the results do not seem so meaningful at the moment

13

When buying a product, I try my best to consider whether its usage will affect the environment or other consumers

14

I believe that some of my low carbon consumption behavior would enhance the quality of the living environment in our surroundings

15

Purchasing carbon-labeled milk products can improve my reputation related to environmental protection

16

I think carbon-labeled milk is safe and reliable to buy

17

Buying low-carbon products is conducive to enhancing the environmental awareness of the entire society

18

I think buying carbon-labeled milk can reduce the CO2 emissions of the dairy industry, thus helping to protect the environment

Risk attitude

Perceived consumer effectiveness

Perceived benefits

(continued)

Appendix

69

(continued) Number

Question item

19

I think people around me who see that I am buying low- carbon products will be positively affected

Note

Demographics of the questionnaire sample. Indicator Gender Age

Number in the sample

Percentage (%)

Cumulative percentage (%)

Female

223

49.2

49.2

Male

230

50.8

100

18–25

247

54.5

54.5

26–35

127

28.0

82.6

36–45

55

12.1

94.7

46–55

16

3.5

98.2

56 or above Level of education

Occupation

8

1.8

100

35

7.7

7.7

High school, 144 vocational middle school, and advanced vocational training

31.8

39.5

College and undergraduate

256

56.5

96.0

Postgraduate or above

18

4.0

100

Student

117

25.8

25.8

Freelancer or self-employed

171

37.7

63.6

Teacher, doctor, scientific staff (etc.)

43

9.5

73.1

Civil servant or public officer

23

5.1

78.1

Business employee

96

21.2

99.3

3

0.7

100

1500 RMB or less

114

25.2

25.2

1500–3000 RMB

120

26.5

51.7

3000–4500 RMB

101

22.3

74.0

4500–6000 RMB

61

13.5

87.4

6000 RMB or above

57

12.6

100

Lower middle school or less

Retired Monthly income

70

2 Consumer Behavior Towards Carbon Labeling Scheme

References Adler AA, Doole GJ, Romera AJ, Beukes PC (2015) Managing greenhouse gas emissions in two major dairy regions of New Zealand: a system-level evaluation. Agr Syst 135:1–9 Arslan T, Yilmaz V, Aksoy HK (2012) Structural equation model for environmentally conscious purchasing behavior. Int J Environ Res 6:323–334 Assarut N, Srisuphaolarn P (2012) Determinants of green product purchase intentions: the roles of environmental consciousness and product attributes. ChulalongKom Bus Rev 32:108–122 AzariJafari H, Yahia A, Amor MB (2016) Life cycle assessment of pavements: reviewing research challenges and opportunities. J Clean Prod 112:2187–2197 Baek CY, Lee KM, Park KH (2014) Quantification and control of the greenhouse gas emissions from a dairy cow system. J Clean Prod 70:50–60 Bai Y, Liu Y (2013) An exploration of residents’ low-carbon awareness and behavior in Tianjin, China. Energ Policy 61:1261–1270 Barnard E, Mitra A (2010) A contingent valuation method to measure willingness to pay for eco-label products. Las Vegas, US Bartiaux F (2008) Does environmental information overcome practice compartmentalisation and change consumers’ behaviours? J Clean Prod 16:1170–1180 Beske P, Land A, Seuring S (2014) Sustainable supply chain management practices and dynamic capabilities in the food industry: a critical analysis of the literature. Int J Prod Econ 152:131–143 Biggs EM, Bruce E, Boruff B, Duncan JMA, Horsley J, Pauli N, McNeill K, Neef A, Ogtrop FV, Curnow J, Haworth B, Duce S, Imanari Y (2015) Sustainable development and the water-energyfood nexus: a perspective on livelihoods. Environ Sci Policy 54:389–397 Birgelen MV, Semeijn J, Keicher M (2009) Packaging and proenvironmental consumption behavior: investigating purchase and disposal decisions for beverages. Environ Behav 41:125–146 Bonti-Ankomah S, Yiridoe EK (2006) Organic and conventional food: a literature review of the economics of consumer perceptions and preferences. In: Final Report, Organic Agriculture Centre of Canada Nova Scotia Agricultural College Borges A, Chebat JC, Babin BJ (2010) Does a companion always enhance the shopping experience? J Retail Consum Serv 17:294–299 Borin N, Cerf DC, Krishnan R (2011) Consumer effects of environmental impact in product labeling. J Consum Mark 28:76–86 Bryman A (eds) (2017) Quantitative and qualitative research: further reflections on their integration. In Mixing methods: qualitative and quantitative research. Routledge, London Cai BF, Feng XZ, Chen XM (2012) Transport carbon dixiode emissions and low carbon development. Chemical Industry Press, Beijing (In Chinese) Campbell MC, Goodstein RC (2001) The moderating effect of perceived risk on consumers’ evaluations of product incongruity: preference for the norm. J Consum Res 28:439–449 Castanheira ÉG, Dias AC, Arroja L, Amaro R (2010) The environmental performance of milk production on a typical Portuguese dairy farm. Agr Syst 103:498–507 Census Office of the People’s Republic of China (2010) Population census data of China in 2010. http://www.stats.gov.cn/tjsj/pcsj/rkpc/6rp/indexch.html Accessed 24 Feb 2014 Chekima B, Wafa S, Igau OA, Chekima S, Sondoh SL (2016) Examining green consumerism motivational drivers: does premium price and demographics matter to green purchasing? J Clean Prod 112:3436–3450 Chen TB, Chai LT (2010) Attitude towards the environment and green products: consumers’ perspective. Manag Sci Eng 4:27–39 Chen X, Corson MS (2014) Influence of emission-factor uncertainty and farm-characteristic variability in LCA estimates of environmental impacts of French dairy farms. J Clean Prod 81:150–157 Cherubini F, Bargigli S, Ulgiati S (2009) Life cycle assessment (LCA) of waste management strategies: Landfilling, sorting plant and incineration. Energy 34:2116–2123

References

71

Chomeya R (2010) Quality of psychology test between Likert scale 5 and 6 points. J Soc Sci 6:399–403 Chou L (2013) Study of green consumption behaviour. Tianjin, China (In Chinese) Corrigan JR, Rousu MC (2006) Posted prices and bid affiliation: evidence from experimental auctions. Am J Agr Econ 88:1078–1090 Dholakia UM, Kahn BE, Reeves R, Rindfleisch A, Stewart D, Taylor E (2010) Consumer behavior in a multichannel, multimedia retailing environment. J Interact Mark 24:86–95 D’Souza C, Taghian M, Lamb P (2006) An empirical study on the influence of environmental labels on consumers. Corp Commun Int J 11:162–173 DEFRA (Department of Environment, Food and Rural Affairs) (2012) Guidelines to Defra/DECC’s GHG Conversion Factors for Company Reporting. https://www.gov.uk/government/publicati ons/2012-guidelines-to-defra-decc-s-ghg-conversion-factors-for-company-reporting-method ology-paper-for-emission-factors. Accessed 20 Feb 2014 Do Paço AMF, Raposo MLB, Filho WL (2009) Identifying the green consumer: a segmentation study. J Target Meas Anal Mark 17:17–25 Dolnicar S, Grün B, Leisch F (2016) Increasing sample size compensates for data problems in segmentation studies. J Bus Res 69:992–999 Dong Y, Xia B, Chen W (2014) Carbon footprint of urban areas: an analysis based on emission sources account model. Environ Sci Policy 44:181–189 Dono J, Webb J, Richardson B (2010) The relationship between environmental activism, proenvironmental behaviour and social identity. J Environ Psychol 30:178–186 Dutreuil M, Wattiaux M, Hardie CA, Cabrera VE (2014) Feeding strategies and manure management for cost-effective mitigation of greenhouse gas emissions from dairy farms in Wisconsin. J Dairy Sci 97:5904–5917 Echeverría R, Hugo MV, Sepúlveda C (2014) Willingness to pay for carbon footprint on foods. Brit Food J 116:186–196 Edwards-Jones G, Plassmann K, York EH, Hounsome B, Jones DL, i Canals LM (2009) Vulnerability of exporting nations to the development of a carbon label in the United Kingdom. Environ Sci Policy 12:479–490 Emanuel R, Adams JN (2011) College students’ perceptions of campus sustainability. Int J Sust Higher Edu 12:79–92 Essoussi LH, Linton JD (2010) New or recycled products: how much are consumers willing to pay? J Consum Mark 27:458–468 Fisher C, Bashyal S, Bachman B (2012) Demographic impacts on environmentally friendly purchase behaviors. J Target Meas Anal Mark 20:172–184 Frazier CB, Ludwig TD, Whitaker B, Roberts DS (2013) A hierarchical factor analysis of a safety culture survey. J Saf Res 45:15–28 Fryxell GE, Lo CWH (2003) The influence of environmental knowledge and values on managerial behaviours on behalf of the environment: an empirical examination of managers in China. J Bus Ethics 46:45–69 Gadema Z, Oglethorpe D (2011) The use and usefulness of carbon labelling food: a policy perspective from a survey of UK supermarket shoppers. Food Policy 36:815–822 Gadenne D, Sharma B, Kerr D, Smith T (2011) The influence of consumers’ environmental beliefs and attitudes on energy saving behaviours. Energy Policy 39:7684–7694 Gao HX (2009) Research on consumers’ perceived risk and behavior patterns. Beijing, China (In Chinese) Ghaffarzadegan N, Lyneis J, Richardson GP (2011) How small system dynamics models can help the public policy process. Syst Dynam Rev 27:22–44 González-García S, Castanheira ÉG, Dias AC, Arroja L (2013) Environmental life cycle assessment of a dairy product: the yoghurt. Int J Life Cycle Ass 18:796–811 Gössling S, Buckley R (2016) Carbon labels in tourism: persuasive communication? J Clean Prod 111:358–369

72

2 Consumer Behavior Towards Carbon Labeling Scheme

Grunert KG, Hieke S, Wills J (2014) Sustainability labels on food products: consumer motivation, understanding and use. Food Policy 44:177–189 Guenther M, Saunders CM, Tait PR (2012) Carbon labeling and consumer attitudes. Carbon Manag 3:445–455 Ha N, Feike T, Back H, Xiao H, Bahrs E (2015) The effect of simple nitrogen fertilizer recommendation strategies on product carbon footprint and gross margin of wheat and maize production in the North China Plain. J Environ Manage 163:146–154 Hagemann M, Ndambi A, Hemme T, Latacz-Lohmann U (2012) Contribution of milk production to global greenhouse gas emissions. Environ Sci Pollut R 19:390–402 Hart PM, Dale R (2014) With or without you: The positive and negative influence of retail companions. J Retail Consum Serv 21:780–787 Hartikainen H, Roininen T, Katajajuuri JM, Pulkkinen H (2014) Finnish consumer perceptions of carbon footprints and carbon labelling of food products. J Clean Prod 73:285–293 Hassan Y, Nor MNAM (2013) Understanding consumer decision making towards green electronic products. In: Proceedings of the Kuala Lumpur International Business, Economics and Law Conference, Kuala Lumpur, Malaysia Hatew B, Bannink A, Van Laar H, De Jonge LH, Dijkstra J (2016) Increasing harvest maturity of whole-plant corn silage reduces methane emission of lactating dairy cows. J Dairy Sci 99:354–368 Henderson CM, Beck JT (2011) Palmatier RW. review of the theoretical underpinnings of loyalty programs. J Consum Psychol 21:256–276 Hernandez-Ortega B (2011) The role of post-use trust in the acceptance of a technology: drivers and consequences. Technovation 31:523–538 Hospido A, Moreira MT, Feijoo G (2003) Simplified life cycle assessment of Galician milk production. Int Dairy J 13:783–796 Huang J, Xu CC, Ridoutt BG, Liu JJ, Zhang HL, Chen F, Li Y (2014) Water availability footprint of milk and milk products from large-scale dairy production systems in Northeast China. J Clean Prod 79:91–97 Huysveld S, linden VV, Meester SD, Peiren N, Muylle H, Lauwers L, Dewulf J (2015) Resource use assessment of an agricultural system from a life cycle perspective- a dairy farm as case study. Agr Syst 135:77-89 Inrig T, Amey B, Borthwick C, Beaton D (2012) Validity and reliability of the fear-avoidance beliefs questionnaire (FABQ) in workers with upper extremity injuries. J Occup Rehabil 22:59–70 IPCC (Intergovernmental Panel on Climate Change) (2006) Guidelines for national greenhouse gas inventories. http://www.ipcc-nggip.iges.or.jp/public/2006gl/ Accessed 7 Aug 2020 Izquierdo I, Olea J, Abad FJ (2014) Exploratory factor analysis in validation studies: uses and recommendations. Psicothema 26:395–400 Joonas K (2008) Environmentally friendly products: factors affecting search for information. AIMS Int J Manag 2:165–176 Khushaba RN, Wise C, Kodagoda S, Louviere J, Kahn BE, Townsend C (2013) Consumer neuroscience: Assessing the brain response to marketing stimuli using electroencephalogram (EEG) and eye tracking. Expert Syst Appl 40:3803–3812 Kim YK, Trail G (2010) Constraints and motivators: a new model to explain sport consumer behavior. J Sport Manag 24:190–210 Kinnear TC, Taylor JR, Ahmed SA (1974) Ecologically concerned consumers: who are they? J Mark 38:20–24 Kotler P, Armstrong G (2010) Principles of marketing, 14th ed. Pearson Prentice Hall, Upper Saddle River, US Lam TH (2014) Awareness of eco-labeling of students of higher education in Hong Kong. Master’s thesis, University of Hong Kong. http://dx.doi.org/https://doi.org/10.5353/th_b5334260. Accessed 24 April 2018 Lee MC (2009) Factors influencing the adoption of internet banking: An integration of TAM and TPB with perceived risk and perceived benefit. Electron Commer Res Appl 8:130–141

References

73

Li J, Colombier M (2009) Managing carbon emissions in China through building energy efficiency. J Environ Manag 90:2436–2447 Li Q, Long R, Chen H (2017) Empirical study of the willingness of consumers to purchase lowcarbon products by considering carbon labels: a case study. J Clean Prod 161:1237–1250 Li X, Jensen KL, Clark CD, Lambert DM (2016) Consumer willingness to pay for beef grown using climate friendly production practices. Food Policy 64:93–106 Lin PC, Huang YH (2012) The influence factors on choice behavior regarding green products based on the theory of consumption values. J Clean Prod 22:11–18 Lindsey-Mullikin J, Munger JL (2011) Companion shoppers and the consumer shopping experience. J Relation Mark 10:7–27 Ling CY (2013) Consumers’ purchase intention of green products: an investigation of the drivers and moderating variable. Mark Manage 57:14503–14509 Liu T, Wang Q, Su B (2016) A review of carbon labeling: Standards, implementation, and impact. Renew Sust Energ Rev 53:68–79 Lobb AE, Mazzocchi M, Traill WB (2006) Risk perception and chicken consumption in the avian flu age - a consumer behaviour study on food safety information. In: Proceedings of the American Agricultural Economics Association Annual Meeting, Long Beach, US. Lombardi GV, Berni R, Rocchi B (2017) Environmental friendly food. Choice experiment to assess consumer’s attitude toward “climate neutral” milk: the role of communication. J Clean Prod 142:257–262 Mainieri T, Barnett EG, Valdero TR, Unipan JB, Oskamp S (1997) Green buying: the influence of environmental concern on consumer behavior. J Soc Psychol 137:189–204 Makower J, Pike C (2009) Strategies for the green economy: opportunities and challenges in the new world of business. New York, US Manasakunkit C, Chinda T (2013) The development of a basic dynamic model of household waste recycling. In: Proceedings of the 4th International Conference on Engineering, Project, and Production Management (EPPM) Maniatis P (2016) Investigating factors influencing consumer decision-making while choosing green products. J Clean Prod 132:215–228 Meneses M, Pasqualino J, Castells F (2012) Environmental assessment of the milk life cycle: the effect of packaging selection and the variability of milk production data. J Environ Manage 107:76–83 Meyer R, Liebe U (2010) Are the affluent prepared to pay for the planet? explaining willingness to pay for public and quasi-private environmental goods in Switzerland. Popul Environ 32:42–65 Milfont TL (2007) Psychology of environmental attitudes: a cross-cultural study of their content and structure. Doctoral Dissertation, University of Auckland, New Zealand Milfont TL, Duckitt J (2010) The environmental attitudes inventory: a valid and reliable measure to assess the structure of environmental attitudes. J Environ Psychol 30:80–94 Mitchell VW (1999) Consumer perceived risk: conceptualisations and models. Eur J Mark 33:163– 195 Moser AK (2015) Thinking green, buying green? Drivers of pro-environmental purchasing behavior. J Consum Mark 32:167–175 Mostafa MM (2016) Egyptian consumers’ willingness to pay for carbon-labeled products: a contingent valuation analysis of socio-economic factors. J Clean Prod 135:821–828 Mowen JC, Minor M (2001) Consumer behavior: a framwork. New York, US Mukherjee J, Ray S, Ghosh PB (2013) A system dynamic modeling of carbon cycle from mangrove litter to the adjacent Hooghly estuary. India Ecol Model 252:185–195 Muñoz Cederberg C, Mattsson B (2000) Life cycle assessment of milk production-a comparison of conventional and organic farming. J Clean Prod 8:49–60 National Bureau of Statistics of the People’s Republic of China (2013) Index to explain: people’s livelihood. http://www.stats.gov.cn/tjsj/zbjs/201310/t20131029_449516.html Accessed 13 Aug 2020

74

2 Consumer Behavior Towards Carbon Labeling Scheme

National Bureau of Statistics of the People’s Republic of China (2014) The statistical bulletin of national economic and social development in 2013. http://www.stats.gov.cn/tjsj/zxfb/201402/t20 140224_514970.html Accessed 13 Aug 2020 NDRC (National Development and Reform Commission) (2014) China regional power grid emission factor report. http://qhs.ndrc.gov.cn/qjfzjz/200907/t20090703_289357.html Accessed 13 Aug 2020 Nigri EM, Barros ACD, Rocha SDF, Romeiro Filho E (2014) Assessing environmental impacts using a comparative LCA of industrial and artisanal production processes: “minas cheese” case. Food Sci Tech 34:522–531 Noor NAM, Mat N, Mat NA, Jamaluddin CZ, Salleh HS, Muhammad A (2012) Emerging green product buyers in Malaysia: their profiles and behaviors. In: Proceedings of 3rd international conference on business and economic research, Bandung, Indonesia Ozaki R (2011) Adopting sustainable innovation: what makes consumers sign up to green electricity? Bus Strategy Environ 20:1–17 Ozdemir A, Altural T (2013) A comparative study of frequency ratio, weights of evidence and logistic regression methods for landslide susceptibility mapping: Sultan Mountains, SW Turkey. J Asian Earth Sci 64:180–197 Park YS, Egilmez G, Kucukvar M (2016) Emergy and end-point impact assessment of agricultural and food production in the United States: a supply chain-linked ecologically-based life cycle assessment. Ecol Indic 62:117–137 Peattie K (2010) Green consumption: behavior and norms. Annu Rev Environ Resour 35:195–228 Peschel AO, Grebitus C, Steiner B, Veeman M (2016) How does consumer knowledge affect environmentally sustainable choices? Evidence from a crosscountry latent class analysis of food labels. Appetite 106:78–91 Pillai P, Meghrajani I (2013) Consumer attitude towards eco-friendly goods - a study of electronic products in Ahmedabad city. Int J Eng Res Technol 2:1124–1127 Qing P, Yan FX, Wang MD (2006) Empirical study of consumer behaviour on green vegetables. Issues Agric Econ 6:73–78 (In Chinese) Rezai G, Mohamed Z, Shamsudin MN (2011) Malaysian consumer’s perceptive towards purchasing organically produce vegetable. In: Proceedings of the 2nd international conference on business and economic research, Langkawi, Malaysia Riera FA, Suárez A, Muro C (2013) Nanofiltration of UHT flash cooler condensates from a dairy factory: characterisation and water reuse potential. Desalination 309:52–63 Rijnsoever FJV, Mossel AV, Broecks KPF (2015) Public acceptance of energy technologies: the effects of labeling, time, and heterogeneity in a discrete choice experiment. Renew Sustain Energy Rev 45:817–829 Rowlings DW, Grace PR, Scheer C, Kiese R (2013) Influence of nitrogen fertiliser application and timing on greenhouse gas emissions from a lychee (Litchi chinensis) orchard in humid subtropical Australia. Agr Ecosyst Environ 179:168–178 Ruscio J, Roche B (2012) Determining the number of factors to retain in an exploratory factor analysis using comparison data of known factorial structure. Psychol Assess 24:282–292 Seppälä J (2003) Life cycle impact assessment based on decision analysis (Ph.D. Dissertation). Helsinki University of Technology, Espoo Sharp A, Wheeler M (2013) Reducing householders’ grocery carbon emissions: carbon literacy and carbon label preferences. Australas Mark J 21:240–249 Shuai CM, Ding LP, Zhang YK, Guo Q, Shuai J (2014a) How consumers are willing to pay for low-carbon products? - Result from a carbon- labeling scenario experiment in China. J Clean Prod 83:366–437 Shuai CM, Yang XM, Zhang YK (2014b) Consumer behaviour on low-carbon agri-food purchase: A carbon labelling experimental study in China. Agr Econ 60:133–146 Spaargaren G, van Koppen CSA, Janssen AM, Hendriksen A, Kolfschoten CJ (2013) Consumer responses to the carbon labelling of food: a real life experiment in a canteen practice. Sociol Ruralis 53:432–453

References

75

Steg L, Vlek C (2009) Encouraging pro-environmental behaviour: an integrative review and research agenda. J Environ Psychol 29:309–317 Steiner M, Wiegand N, Eggert A (2016) Platform adoption in system markets: the roles of preference heterogeneity and consumer expectations. Int J Res Mark 33:276–296 Sumari S, Ibrahim R, Zakaria NH, Hamid AHA (2013) Comparing three simulation model using taxonomy: system dynamic simulation, discrete event simulation and agent based simulation. Int J Manag Excell 1:54–59 Tam R, Yassa B, Parker H et al (2016) On campus food purchasing behaviours, preferences and opinions on food availability of university students. Nutrition 73:15 Tan BC (2011) The roles of knowledge, threat, and PCE on green purchase behaviour. Int J Bus Manag 6:14–27 Tehrani AF, Ahrens D (2016) Enhanced predictive models for purchasing in the fashion field by using kernel machine regression equipped with ordinal logistic regression. J Retail Consum Serv 32:131–138 Tonsor GT, Schroeder TC, Pennings JME (2009) Factors impacting food safety risk perceptions. J Agric Econ 60:625–644 Triandis HC (1989) The self and social behavior in differing cultural contexts. Psychol Rev 96:506 Trope Y, Liberman N (2010) Construal-level theory of psychological distance. Psychol Rev 117:440 Upham P, Dendler L, Bleda M (2011) Carbon labelling of grocery products: public perceptions and potential emissions reductions. J Clean Prod 19:348–355 Vanclay JK, Shortiss J, Aulsebrook S, Gillespie AM, Howell BC, Johanni RR, Maher MJ, Mitchell KM, Stewart MD, Yates J (2011) Customer response to carbon labelling of groceries. J Consum Policy 34:153–160 Van-Middelaar CE, Berentsen PBM, Dijkstra J, De Boer IJM (2013) Evaluation of a feeding strategy to reduce greenhouse gas emissions from dairy farming: the level of analysis matters. Agr Syst 121:9–22 Vecchio R, Annunziata A (2015) Willingness-to-pay for sustainability-labelled chocolate: an experimental auction approach. J Clean Prod 86:335–342 Vergé XPC, Maxime D, Dyer JA, Desjardins RL, Arcand Y, Vanderzaag A (2013) Carbon footprint of Canadian dairy products: calculations and issues. J Dairy Sci 96:6091–6104 Virtanen Y, Kurppa S, Saarinen M (2011) Carbon footprint of food-approaches from national input– output statistics and a LCA of a food portion. J Clean Prod 19:1849–1856 Vlachos D, Georgiadis P, Iakovou E (2007) A system dynamics model for dynamic capacity planning of remanufacturing in closed-loop supply chains. Comput Oper Res 34:367–394 Wang X, Kristensen T, Mogensen L, Knudsen MT, Wang X (2016) Greenhouse gas emissions and land use from confinement dairy farms in the Guanzhong plain of China–using a life cycle assessment approach. J Clean Prod 113:577–586 Wärneryd KE (1996) Risk attitudes and risky behavior. J Econ Psychol 17:749–770 Webb DJ, Mohr LA, Harris KE (2008) A re-examination of socially responsible consumption and its measurement. J Bus Res 61:91–98 Weber EU, Blais AR, Betz NE (2002) A domain-specific risk-attitude scale: measuring risk perceptions and risk behaviors. J Behav Decis Mak 15:263–290 Wheeler T, Von Braun J (2013) Climate change impacts on global food security. Science 341:508– 513 Woon KS, Lo IM (2013) Greenhouse gas accounting of the proposed landfill extension and advanced incineration facility for municipal solid waste management in Hong Kong. Sci Total Environ 458:499–507 Wu P, Low SP, Xia B, Zuo J (2014) Achieving transparency in carbon labelling for construction materials-Lessons from current assessment standards and carbon labels. Environ Sci Policy 44:11–25 Xie M, Li L, Qiao Q, Sun Q, Sun T (2011) A comparative study on milk packaging using life cycle assessment: from PA-PE-Al laminate and polyethylene in China. J Clean Prod 19:2100–2106

76

2 Consumer Behavior Towards Carbon Labeling Scheme

Yin SJ (2010) Empirical study of Chinese organic food market based on the perspective of consumer behavior. Doctoral dissertation, Jiangnan University, Jiangsu. (In Chinese) Zhao R, Deutz P, Neighbour G, McGuire M (2012) Carbon emissions intensity ratio: an indicator for an improved carbon labelling scheme. Environ Res Lett 7:9 Zhao HH, Gao Q, Wu YP, Wang Y, Zhu XD (2014) What affects green consumer behavior in China? A case study from Qingdao. J Clean Prod 63:43–151 Zhao R, Zhong S (2015) Carbon labelling influences on consumers’ behaviour: a system dynamics approach. Ecol Indic 51:98–106 Zhao R, Geng Y, Liu Y, Tao X, Xue B (2018a) Consumers’ perception, purchase intention, and willingness to pay for carbon-labeled products: a case study of Chengdu in China. J Clean Prod 171:1664–1671 Zhao R, Xu Y, Wen X, Zhang N, Cai J (2018b) Carbon footprint assessment for a local branded pure milk product: A lifecycle based approach. Food Sci Technol 38:98–105 Zsóka AN (2008) Consistency and “awareness gaps” in the environmental behaviour of Hungarian companies. J Clean Prod 16:322–329

Chapter 3

Interaction Among Stakeholders Involved in Carbon Labeling Scheme

Abstract The implementation of a carbon labeling scheme provides an opportunity to drive enterprises have a holistic assessment on the lifecycle environmental impact of their services or products, which also encourages producers to pay more attentions to the corporate social responsibility. However, the carbon labeling practice is voluntary for enterprises. Government plays a leading role in developing welldesigned policies to transform industrial innovation into product sustainability and thus promote sustainable performance. Their interactions are complex, which may decide whether the labeling scheme will successfully be implemented or not. This chapter applies game theory, combined with system dynamics to modeling strategic interaction between enterprises, consumers and government in a carbon-labeled product market. Keywords Game theory · System dynamics · Stakeholders · Interaction

3.1 Introduction With the development of low carbon economy, the carbon labelling scheme has become a novel mean for product communication to promote the green consumption (Guenther et al. 2012). Green consumption refers to a sustainable mode of consumption, i.e., environmental purchasing, where consumers intentionally select products that have lower pollution and energy consumption, to satisfy their purchasing demands as well as achieve environmental protection (OECD 2016). Their preferences on products’ green performance, e.g., demand of carbon labeled products or services, may partially result in the enterprises’ actions on fostering a low-carbon production system. The implementation of a carbon labelling scheme provides an opportunity to drive enterprises have a holistic assessment on the lifecycle environmental impact of their services or products (Zhao et al. 2017). It can also encourage producers to pay more attentions on the corporate social responsibility. Such a system may facilitate producers to support more innovative activities (such as eco-design, cleaner production, green marketing, etc.) on developing environmentally friendly products and guiding the consumers to change their behaviors toward low carbon consumption (Dangelico and Pontrandolfo 2010). © Springer Nature Singapore Pte Ltd. 2021 R. Zhao and Y. Geng, Carbon Labeling Practice, https://doi.org/10.1007/978-981-16-2583-1_3

77

78

3 Interaction Among Stakeholders Involved in Carbon Labeling Scheme

However, it is difficult to visualize the performance of these policies, due to the complexity of a sustainable operation for all the participants, e.g., trustworthy and loyal partnership (Myeong et al. 2014; Choi 2014, 2015). Consumers tend only to accept lower price premiums, because their willingness to pay for the carbon-labeled products are significantly affected by individual characteristics such as age, gender, income, and education level (Ramayah et al. 2010; Olive et al. 2011). For enterprises, the additional cost of low-carbon technologies, market risk, and complexity of the external business environment may result in uncertainty regarding commercial success (Zhao et al. 2013; Shuai et al. 2014; Bi et al. 2015). Though the enterprises’ involvement is often voluntary, they usually respond to a range of signals (e.g., consumer demand, governmental policies). For this reason, governments play a leading role in developing well- designed policies to drive industrial innovation into product sustainability and promote sustainable performance (Kanada et al. 2013; Choi 2015). Their interactions are complex, which may decide whether the scheme will successfully be implemented or not. Given this background, this chapter applies game theory, combined with system dynamics (SD) to modeling strategic interaction between enterprises, consumers and government in a carbon-labeled product market. Game theory can predict each player’s optimal strategy when conflicts occur between groups (Hui and Bao 2013). A game usually makes a unique prediction from the possible strategic actions that each player may choose, and this prediction is indicated by an equilibrium state, i.e., the Nash equilibrium (Zhao et al. 2016). However, the process to seek for such an equilibrium state is difficult to be investigated by game theory application, e.g., when to have transient transformation in a game (Kim and Kim 1997). System dynamics is useful to bridge this gap by visual simulation to help decision makers better understand the complex and dynamic process of a game evolution (Yunna et al. 2015). Their combination allows for the full utilization of their advantages, to identify the strategic evolution of the stakeholders, to seek for the market development of carbon labelled products. The chapter is further divided into three sections. Section 3.1 applies a game theoretical analysis combined with system dynamics to model strategic interaction between enterprises and consumers with bounded rationality in a carbon labelled product market. Section 3.2 presents an evolutionary game model to examine how enterprises respond to a range of governmental incentive policies related to the implementation of a carbon labeling scheme, namely a direct subsidy and a series of preferential tax rates. SD is then employed to simulate the created game model by using scenario analysis based on two scenarios: an individual and a combined policy intervention. Section 3.3 proposes a SD approach to simulate enterprises’ response to a combined government policy instrument, that is, subsidies and penalties, in the implementation of a carbon reduction labelling scheme.

3.2 Fundamentals of Game Theory

79

3.2 Fundamentals of Game Theory In this Section, a well-known game example entitled “The Prisoner’s Dilemma” (Davis 1983; Luce and Raiffa 1989; Straffin 1993) is introduced to demonstrate the fundamentals of game theory. The significance of Prisoner’s Dilemma is that considerable social phenomena or managerial issues can be abstracted as the prisoner’s choices (Straffin 1993). Such classical game can be seen as a prototype to develop game scenarios regarding interactions among stakeholders involved in carbon labeling scheme, which will be presented in details in Sects. 3.3, 3.4 and 3.5. For instance, the carbon labelled products are advocated as an assumption in our study, which may lead to further competition within those enterprises that produce the same type of products. It is assumed that enterprise will not only engage in a cost reduction war, but will also become involved in a product quality improvement, e.g. reducing the carbon footprint associated with products. If both of the enterprises cut the economic cost of production, and lay stress on environmental quality of the products, they may have the opportunity to explore a new market share. Otherwise, if both of the enterprises do not put emphasis on the environmental consideration, but just maximize their own economic interests, their long term sales profit may decrease gradually. If one enterprise adopts cleaner production to save the manufacturing cost and energy and provides carbon labelled product, while another enterprise still insists on their inherent action for production without taking any environmental considerations into account, the former enterprise apparently may have advantages to attract more customers to buy their products, thus increasing profits. The Prisoner’s game is originally generated by Albert W. Thucker (Romp 1997). There are two players in the game, who are detained in separate rooms as being the two suspect criminals in a particular case. The policeman is very experienced, but without adequate evidence to sentence them in the trial process. Therefore, he tries to give each player two alternative choices, to confess what they have done, or not to confess. Therefore, the dilemma faced by the two prisoners is to choose whether to confess or not. The potential combinations of the choices that they would make are presented as follows. If both of them decide not to confess, due to the absence of direct evidence of their guilt, they may be imposed in a lighter punishment, i.e. both of them are sentenced to one year in jail. If both of them decide to confess, they will be found guilty and sentenced to less severe punishment. In such case, both of them will be sentenced to six years in jail. If one of them determines to confess and another one rejects, the confessor will receive a reduced penalty as a reward, whilst his partner will be punished heavily. It is assumed the confessor can be exempt from jail, but his partner who denied confessing, will be sentenced by ten years. Here, the penalties can be arbitrary, and only seen to be illustrative in this example. According to the above set

80

3 Interaction Among Stakeholders Involved in Carbon Labeling Scheme

Table 3.1 Payoffs matrix of the “Prisoner’s dilemma”

Prisoner 1

Prisoner 2 Not confess

Confess

Not confess

(−1, −1)

(−10, 0)

Confess

(0, −10)

(−6, −6)

of potential strategic actions, this dilemma can be described in a matrix, shown in Table 3.1. The “prisoner’s dilemma” is a typical type of the “strategic form game”. If a game is defined as the strategic form game, the game should contain the following three important factors, which are “the list of players”, “the set of strategies available to each player”, “the payoffs associated with any strategy combination (one strategy per player)”, respectively (Dutta 1999). Payoffs in game theory can be described as the real numbers “positive or negative” given for different situations. These numbers are generally derived by the outcomes in the game theory. In some cases, the payoff is using a logic unit instead of the monetary value. For instance, if the outcome can be interpreted as win, lose and draw, accordingly there are values “1”, “−1”, “0” to stand for these situations separately (Rapoport 1999; Thomas 2003). Within the “strategic form game”, the simplest form is the “two person game”, which describes that two players are in the game situation and each of them has two available strategies for further decision making. Moreover, the two person game can be divided into two categories, within which one is called “two-person zero-sum game” and another one is called “two- person non zero-sum game” (Rapoport 1999). A “two-person zero-sum” game is a game that the player’s payoffs add up to be zero no matter what strategy they would use (Kelly 2003; Thomas 2003). In such a game situation, the competition is very strict, that one player will win and the other will lose. The “two-person zero-sum” game is applicable for daily life, but in some real circumstance, the payoffs cannot equal to zero as the game is unfair to some extent. However, it is found that the sum could be a constant as the result of the extreme competitive game. In order for simplicity, the term “zero-sum” can be interpreted as “constant sum” in a way which means “players have diametrically opposed the interests” (Davis 1983). The “Prisoner’s dilemma” is a typical “two person non-zero-sum game” by the reason that the total payoff is twelve years if both players decide to confess and two year if they do not confess. In contrast with the two person zero-sum game, it is a non-strictly competitive game in which the players are not “completely antagonistic to one another” (Luce and Raiffa 1989; Thomas 2003). In the non zero-sum game, the outcome could be beneficial and acceptable to all the players involved, as no one will win everything but everyone will get something instead. To some extent, this can be understood as a “win–win” situation which is different from the zero-sum game as strict “win-lose” situation. Many conflicts related to economic, political and military interests can be transformed into a non zero-sum game situation. For example, the game scenarios of medical products development could be based on

3.2 Fundamentals of Game Theory

81

the non zero-sum game with various degrees of cooperation and competition (Luce and Raiffa 1989).

3.2.1 Solution by Dominated Strategy Elimination The solution of “two person non zero-sum game”, e.g. “Prisoners’ Dilemma”, is determined by identification of the “Nash Equilibrium”. Assume that a game makes the unique prediction from the possible strategic actions that each player may choose. The predicted strategy should be a best response compared with the strategies chosen by all the other players. This Section presents how to find out the Nash Equilibrium by using the method of dominated strategy elimination (Romp 1997; Gintis 2009). In applying this method, each player should be examined in turn and all those strategies that are strictly dominated should be eliminated. This process may rule out all but one strategy left for each player. Therefore, this method provides a unique solution for the game. Table 3.2 shows the different payoffs between two prisoners while taking the different strategic actions, “Confess” or “not confess”. First, identify the optimum strategy for Prisoner 1, dependent upon what the prisoner 2 may choose. Suppose Prisoner 1 predicts that Prisoner 2 may choose “Confess”,  the  best strategy for Prisoner 1 is also to “confess” derived from the −6 matrix , since −6 is better than −10. As the payoffs are used to represent the −10 duration of imprisonment, the numerical values should be as least as possible. Thus, it is shown in Table 3.3 by underlining and overstriking the first payoff element “−6” when the Prisoner 2 chooses “confess”. If the Prisoner 1 predicts that Prisoner 2 may choose the strategic action “Not confess”,   the best strategy for Prisoner 1 is also to “confess” derived from the matrix 0 , as 0 is better than −1. In this case, Prisoner 1 could be exempt from jail as all −1 Table 3.2 Prisoners’ dilemma in normal form

Table 3.3 Prisoner 1’s optimal choice while Prisoner 2 selecting “confess”

Prisoner 1

Prisoner 1

Prisoner 2 Confess

Not confess

Confess

−6, −6

0, −10

Not confess

−10, 0

−1, −1

Prisoner 2 Confess

Not confess

Confess

−6, −6

0, −10

Not confess

−10, 0

−1, −1

82

3 Interaction Among Stakeholders Involved in Carbon Labeling Scheme

Table 3.4 Prisoner 1’s optimal choice while Prisoner 2 selecting “Not confess”

Prisoner 1

Prisoner 2 Confess

Not confess

Confess

−6, −6

0, −10

Not confess

−10, 0

−1, −1

the punishment will be undertaken by Prisoner 2. Thus, it is considered to underline and overstrike “0” in Table 3.4. Similarly, determine the optimal strategy for Prisoner 2, by forecasting the possible action from Prisoner 1 (taking the array into account). Suppose that Prisoner 1 may choose “confess”, and the best strategy for Prisoner 2 is also to “confess” derived from the matrix [−6, −10], as −6 is also better than −10. Thus, it is considered to underline and overstrike “−6” in Table 3.5. When Prisoner 1 may select “not confess”, Prisoner 2 would better to choose “confess”. From the matrix [0, −1], Prisoner 2 could be exempted from the jail. Accordingly, the selected payoff element is shown in Table 3.6. In summary, all the selected optimal payoffs are shown in Table 3.7, with being underlined and over-striked. If the payoffs in the same box are underlined and overstriked, it is deemed that the corresponding strategic pair is the dominant strategy in the game. In the prisoners’ dilemma, it is clear that there is only one box where both payoffs have been underlined and over-striked, which corresponds to both prisoners confessing. Thus, the Nash equilibrium is unique in this game. Table 3.5 Prisoner 2’s optimal choice while Prisoner 1 selecting “confess”

Table 3.6 Prisoner 2’s optimal choice while Prisoner 1 selecting “Not confess”s

Table 3.7 Prisoner 2’s optimal choice while Prisoner 1 selecting “Not confess”

Prisoner 1

Prisoner 1

Prisoner 1

Prisoner 2 Confess

Not confess

Confess

−6, −6

0, −10

Not confess

−10, 0

−1, −1

Prisoner 2 Confess

Not confess

Confess

−6, −6

0, −10

Not confess

−10, 0

−1, −1

Confess

Not confess

Confess

−6, −6

0, −10

Not confess

−10, 0

−1, −1

Prisoner 2

3.2 Fundamentals of Game Theory

83

However, this result seems to be not a ‘satisfactory’ solution as it leads the payoffs pair (−6, −6), which is worse than (−1, −1), if both players choose “confess”. This is because the individual rationalism could be superior to the group rationality in the context of Non-cooperative game theory. It assumes that the individual action is determined to only act in self-interest (Romp 1997; Thomas 2003). In contrast, the players could enter into a binding and enforceable agreement in the field of the cooperative game, so that the two prisoners are better to choose “not confess”. As individuals are assumed to work together in a voluntary instead of a compulsory way, the non-cooperative game theory research is more prevalent in the economic activities (Romp 1997). Thus, the following studies on application of game theory to analysis of stakeholders’ interactions are all based upon this premise, and affiliated with the non-cooperative game.

3.2.2 Solution by Graphic Method In the premise of “Nash Theorem”, any two-person game (zero-sum or non zero-sum) with a finite number of pure strategies has at least one equilibrium pair. Suppose a pair of strategies x ∗ ∈ X , y ∗ ∈ Y is an equilibrium pair for a non zero-sum game, the necessary and sufficient conditions should be satisfied for any x ∈ X, y ∈ Y :     e1 x ∗ , y ∗ ≥ e1 x, y ∗

(3.1)

    e2 x ∗ , y ∗ ≥ e2 x ∗ , y

(3.2)

The “Prisoner’s dilemma” game still uses “Nash Theorem” to find out all the equilibrium pairs. In order for computation, the normal form of “Prisoner’s dilemma” (See Table 3.2) can be expressed in a matrix form below. 

(−6, −6)(0, −10) (−10, 0)(−1, −1)



In this example, suppose the mix-strategies for prisoner I and II are (x, 1 − x), and (y, 1 − y), respectively. According to the inequality (3.1), for a particular y which is part of an equilibrium pair, x can be found to maximize e1 (x, y). Thus, the x must be its partner in it. If x = 1 which means no matter what y is, prisoner 1 should choose “confess” as the pure strategy. Thus, the expected payoffs of prisoner 1 should be less than or equal to the payoffs once “confess” has been selected. The expression can be obtained as follows: e1 (x, y) ≥ e1 (1, y) where e1 (x, y) can be expressed as followed:

(3.3)

84

3 Interaction Among Stakeholders Involved in Carbon Labeling Scheme

e1 (x, y) = −6x y + 0x(1 − y) − 10(1 − x)y − 1(1 − x)(1 − y) = 3y(x − 3) + (x − 1)

(3.4)

Let x = 1, and substitute the numerical value into Eq. (3.4) and the inequality (3.3), the expressions with indeterminate form can be transformed as follows: 3y(x − 3) + (x − 1) ≥ −6y

(3.5)

Through further mathematical simplification as the merger of similar items for the above expressions, the final expressions can be obtained as below:   1 ≥0 3(x − 1) y + 3

(3.6)

Followed by the constraint conditions of x and y, the following linear simultaneous inequalities (3.7) can be solved if and only if x = 1.   ⎧ 1 ⎨ 3(x − 1) y + 3 ≥ 0 0≤x ≤1 ⎩ 0≤y≤1

(3.7)

Similarly, any fixed x, y could be found to maximise e2 (x, y). If x is part of an equilibrium pair, y should be its partner. If y = 1 which means no matter what x is, prisoner 2 will choose “confess” as the pure strategy. Thus, the expected payoff of prisoner 2 should be less than or equal to the payoff once “confess” being selected. Thus, the expression should satisfy the following: e2 (x, y) ≥ e2 (x, 1)

(3.8)

where e2 (x, y) can be expressed as followed: e2 (x, y) = −6x y − 10x(1 − y) + 0(1 − x)y − 1(1 − x)(1 − y) = 3x(y − 3) + (y − 1)

(3.9)

Let y = 1 and substitute the value into Eq. (3.9) and the inequality (3.8), thus the expressions with indeterminate form can be derived as follows: 3x(3 − y) + (1 − y) ≥ −6x

(3.10)

The simplified expressions can be obtained as followed, by means of the merger of similar items from the above inequality (3.10). 

1 3(y − 1) x + 3

 ≥0

(3.11)

3.2 Fundamentals of Game Theory

85

Fig. 3.1 The equilibrium pair for prisoners’ dilemma

(1, 1) 1

0 1 Combining the constraint conditions of x and y with the inequality (3.11), the following linear simultaneous inequalities (3.12) can be solved if and only if y = 1. ⎧   ⎪ 3(y − 1) x + 13 ≥ 0 ⎪ ⎪ ⎪ ⎪ ⎨ 0≤x ≤1 ⎪ ⎪ ⎪ ⎪ ⎪ ⎩ 0≤y≤1

(3.12)

Thus, (x, y) = (1, 1) should be the unique Nash equilibrium solved by the inequalities (3.7) and (3.12), where the corresponding payoffs are (−6, −6) that means each prisoner should be sentenced to 6 years in jail. The equilibrium pair is shown in Fig. 3.1. Moreover, the Nash equilibrium reflects the probability that both players determine to confess is one, whilst the probability of not confessing is zero. As the “Prisoner’s dilemma” is a pure-strategic game, it shows strictly dominant strategy. Both of the two methods show that the optimum strategy is both to confess in the context of “Non-cooperative Game”. However, the “prisoner’s dilemma” is used to express the simplest situation of a game, the real game among stakeholders involved in the carbon labeled product market is more complex, which is presented in the following Sections.

3.2.3 Summary A classical game entitled “Prisoners’ dilemma” is introduced as an example to illustrate how game theory is applied to decision making. In particular, any game to be built further in the study may consider the “Prisoner’s Dilemma” as an analogy. The game problem can be generally solved by identification of the Nash

86

3 Interaction Among Stakeholders Involved in Carbon Labeling Scheme

Equilibrium, to predict strategic action for all the involved players. Two basic approaches as “elimination of dominated strategy” and “graphic method” are presented to find out the Nash equilibrium. Following Sections describe games in promotion of carbon labelling scheme, and discusses the possible strategic options of the involved players: consumer, enterprise and government, by showing how game theory could be applied to better understand their interactions.

3.3 A Game Between Consumers and Enterprises In development of carbon labeled product market, enterprises and consumers play equally important roles, and their decision-makings may affect one another. This section applies game theoretical analysis to illustrating the interactions of the two players, and modelling possible development of the market for carbon labelled products. It aims to provide a theoretical basis for low-carbon production, and improve consumers’ green consumption awareness, thus to promote sustainable development. As a classical game is generally limited by its assumption on complete rationality for involved players, this study constructs an evolutionary game to investigate the transformation of the player’s strategic actions, to identify the appropriate strategic action for low-carbon consumption and production. Evolutionary game theory focuses on the interactions among conflicted players whose strategic behaviors have direct impact on their payoffs (Zhao et al. 2016). The essence of the evolutionary game theory is to seek for the frequencies of a strategic action taken by the players as the criterion for decision-making in the game process (Ji et al. 2015). It can further describe the complex relationship between change of strategic action and the corresponding payoff fluctuation, which is a powerful tool to reflect cooperation or competition among the players (Tian et al. 2014). For this advantage, it has been widely applied to a number of research fields, such as technological innovation, power management, supply chain management, and resource allocation. SD can help decision makers improve their understanding of the complex feedback structure of a system (Kreng and Wang 2013; Yunna et al. 2015). A game makes a unique prediction from the possible strategic actions that each player may choose. However, this solution is ultimately indicated by an equilibrium state (i.e., the Nash equilibrium), and the involved dynamic and transient transformation is often neglected (Kim and Kim 1997). SD bridges this gap by simulating the embodied game scenarios visually in terms of their non-linear feature (Suryani et al. 2010).

3.3.1 Game Theoretical Model This study considers a game composed of two players: the consumers and the enterprises. Both of them are hypothesized as economic men with bounded rationality, who determine their optimal strategic actions through continuous comparisons of

3.3 A Game Between Consumers and Enterprises

87

their gains and losses. The enterprises have two strategic options: one is to apply carbon labeling, that is to mark the life-cycle-based carbon emissions on their products through labeling, in order to seek opportunities on energy savings and emissions reduction, while the second is not to apply any carbon label. Whether the enterprises decide to implement carbon labeling or not are influenced by cost, economic benefit, and policy incentives. Consumers also have two strategic options: one is to buy carbon-labeled products; the second is not to buy. The key factors that consumers consider when making purchases are their expenditure and product quality. Based on the above strategic choices, the study proposes five hypotheses for the model construction. It should be noted that all the parameters listed in the following hypotheses are non-negative. H1: When not applying carbon label to products, the enterprises’ unit production cost is Ca, and unit sale price is Pa. When applying carbon label, the enterprises’ unit production cost is changed as Cb, and unit sale price is Pb. This satisfies equations Cb > Ca, Pb ≥ Pa. The only differences between these two products are the cost and price. H2: In order to incentivize implementation of the carbon labelling scheme, government provides a subsidy for per-unit production cost Sm to the enterprises that take certified carbon labelling practice. H3: Direct and indirect benefits may be generated from the purchase of carbon labelled products by consumers. Herein, direct benefits contain the per-unit price subsidy offered by the government, Sc. Indirect benefits include difference between the per-unit product psychological price (Pw), and the real price (P): Pw – P. H4: Consumers’ psychological price is deemed as their willingness to pay, i.e., the price that consumers are willing to pay for a product, which is determined by their perceptional judgement (Johnstone and Tan 2015). This study assumes that the per-unit psychological price Pw is affected by three factors: (1) the product’s value in use, g ∈ [0, 1], indicated the degree to which the product meets the needs of the consumer, which has no relationship with carbon label; (2) The product’s green perception, E ∈ [0, 10], represented the consumers’ subjective judgement of the product’s greenness, which is affected by consumers’ environmental knowledge, awareness, education level etc. (Tan et al. 2016). When E = 0, this shows that the product has not included a carbon label. (3) The consumers’ perception of risk, R ∈ [0, 10], which can be understood as the consumers’ trust of the product, or their subjective judgment incurred by purchasing the carbon labelled product, which is affected by the consumers’ purchasing experience, product information, brand reputation etc. (Zhao and Zhong 2015). Perceived risk has a negative effect on the psychological price. For products that have not included carbon label, R = 0. The per-unit psychological cost of consumers can be expressed as: Pw = (g + E − R)k

(3.13)

Herein, k refers to the comprehensive payment, that is, the overall per-unit price that the consumer is willing to pay for the product.

88

3 Interaction Among Stakeholders Involved in Carbon Labeling Scheme

Table 3.8 Evolutionary game pay-off matrix for the enterprises and consumers Consumers Enter-prises

Apply Carbon label (θ)

Buy (γ ) Sm + Pb − Cb,

Not Buy (1 − γ ) (Sm − Cb − Rk)

Sc + (g + E − R)k − Pb Not apply (1 − θ)

(Pa − Ca, gk − Pa)

(−Ca, 0)

H5: The enterprises’ probability of taking strategic action L1 (applying carbon label) is θ. The probability of the enterprise taking strategic action L2 (not applying carbon label) is 1 − θ, where 0 ≤ θ ≤ 1. The consumers’ probability of taking strategic action B1 (buy) is γ , and their probability of taking strategic action B2 (not buy) is 1 − γ , where 0 ≤ γ ≤ h. The pay-off matrix for the enterprises and consumers is given in Table 3.8. According to Table 3.8, the expected pay-offs when the enterprise chooses to “apply carbon label” and to “not apply carbon label” are E L1 and E L2 , with the average pay-off of E L , which are measured as follows: E L1 = γ (Sm + Pb − Cb) + (1 − γ )(Sm − Cb) = γ Pb + Sm − Cb E L2 = γ (Pa − Ca) + (1 − γ )(−Ca) = γ Pa − Ca

(3.14) (3.15)

E L = θ E L1 + (1 − θ)E L2 = θ γ (Pb − Pa) + θ (Sm + Ca − Cb) + γ Pa − Ca (3.16) From Eqs. (3.14), (3.15), and (3.16), the duplicative dynamic equation of enterprises’ strategic action is obtained: F(θ ) =

  dθ = θ (E L1 − E L ) = θ (1 − θ ) γ (Pb − Pa) + Sm + Ca − Cb (3.17) dt

Similarly, the consumers’ expected pay-offs when choosing to “buy” and to “not buy” are EB1 , EB2 , with an average pay-off of EB , which are given as follows: E B1 = θ [Sc + (g + E − R)k − Pb] + (1 − θ)(gk − Pa) = θ [Sc + (E − R)k − Pb + Pa] + gk − Pa

(3.18)

E B2 = θ (−Rk) + (1 − θ) · 0 = −θ Rk

(3.19)

3.3 A Game Between Consumers and Enterprises

89

rate of buying no labelled consumer of buying

comsumer of no buying

rate of buying labelled

probability of consumer buying

R difference of EB1 EB2

L2B1 Pb B1L1

k

Sc

L1B1

EB1 EL1

B2L1 g EB2

E

Sm

Cb

B1L2

Ca

EL2

L1B2

B2L2

L2B2 difference of EL1 EL2

probability of enterprise labelling enterprise of labelling

rate of labelling

enterprise of no labelling

Pa

rate of no labelling

Fig. 3.2 SD Model for the evolutionary game between the enterprises and consumers

E B = γ E B1 + (1 − γ )E B2 = γ θ (Sc + Ek − Pb + Pa) + γ (gk − Pa) − θ Rk (3.20) From Eqs. (3.17), (3.18), and (3.19), the duplicative equation of consumers’ strategic action is obtained: F(γ ) =

dγ = γ (E B1 − E B ) = γ (1 − γ )[θ (Sc + Ek − Pb + Pa) + gk − Pa] dt (3.21)

According to the proposed game, this paper further constructs a SD model by using Stella 9.1.4 to investigate the behavioral variations of the two players, shown in Fig. 3.2. The external variables of the SD model are consistent with those established in the game model hypotheses, whilst other variables are given in Table 3.9.

3.3.2 Theoretical Analysis for Evolutionary Stability In this section, stability analysis for the strategic actions of the enterprises and consumers are conducted respectively, in order to set the conditions for both players

90

3 Interaction Among Stakeholders Involved in Carbon Labeling Scheme

Table 3.9 List of variables for the SD model Variable

Interpretation

Variable

Interpretation

B1L1

Pay-off to consumer purchase when the enterprise applies a carbon label

L1B1

Pay-off to the enterprise applying carbon label when the consumer purchases

B1L2

Pay-off to consumer purchase when the enterprise does not apply a carbon label

L1B2

Pay-off to the enterprise applying carbon label when the consumer does not purchase

B2L1

Pay-off to consumer not purchasing when the enterprise applies a carbon label

L2B1

Pay-off to the enterprise not applying carbon label when the consumer purchases

B2L2

Pay-off to consumer not purchasing when the enterprise does not apply a carbon label

L2B2

Pay-off to the enterprise not applying carbon label when the consumer does not purchase

EB1

Expected pay-off for consumer while choosing to purchase

EL1

Expected pay-off to the enterprise when applying carbon label

EB2

Expected pay-off to consumer while choosing not to purchase

EL2

Expected pay-off to the enterprise when not applying carbon label

Variable

Interpretation

Difference of EB1 EB2

Difference between the expected pay-offs for the consumer’ actions of purchasing and not purchasing

Difference of EL1 EL2

Difference between the expected pay-offs for the enterprise’s actions of applying carbon label and not applying carbon label

Probability of consumer buying

The probability of the consumer choosing the “purchase” action: γ

Probability of enterprise labelling

The probability of the enterprise choosing the “labelling” action: θ

Rate of buying labelled

The rate of change of consumer taking the “purchase” action: dγ /dt

Rate of labelling

The rate of change of enterprise taking the “labelling” action:dθ/dt

3.3 A Game Between Consumers and Enterprises

91

generating mixed strategies into context. A mixed-strategy is that at least one player involved in the game randomizes over some or all of the pure strategies, i.e. a probabilistic distribution is placed by the strategies selection (Gintis 2009). It may avoid the issue related to a defined game result caused by pure strategy, indicating a strong dominance in player’s preferred strategic action, e.g., the player firmly chooses the action “buying carbon labelled product”. , then F(θ ) = 0, indicating that each θ is stable. From Eq. (3.17), if γ = Cb−Ca−Sm Pb−Pa Cb−Ca−Sm If γ = Pb−Pa , then F(θ ) = 0, indicating that there are two points of stability at θ = 0 and θ = 1. The evolutionary stable strategy (ESS) requires F  (θ ) < 0, and the derivative of F(θ ) can be expressed as follows: F (θ ) =

  dF(θ ) = (1 − 2θ ) γ (Pb − Pa) + Sm + Ca − Cb dθ

(3.22)

Pb ≥ Pa has been proposed by H1, through which a number of scenarios are divided: 1.

If Pb = Pa: ➀ ➁

2.

  When Sm + Ca − Cb > 0, F  (θ )θ=0 > 0, and F  (θ )θ=1 < 0, θ = 1 is the ESS;   When Sm + Ca − Cb < 0, F  (θ )θ=0 < 0, and F  (θ )θ=1 > 0, θ = 0 is the ESS.

if Pb > Pa: ➀ ➁ ➂

When Cb − Ca − Sm < 0, and Cb−Ca−Sm < 0, with the constant γ > Pb−Pa Cb−Ca−Sm , the ESS is θ = 1; Pb−Pa > 1, with the When Cb − Ca − Sm > Pb − Pa > 0, and Cb−Ca−Sm Pb−Pa , the ESS is θ = 0; constant γ < Cb−Ca−Sm Pb−Pa When Pb − Pa > Cb − Ca   − Sm > 0, and when γ > Cb−Ca−Sm     F > 0, F < 0, the ESS is θ = 1; when (θ ) (θ ) Pb−Pa θ=0  θ=1  Cb−Ca−Sm    γ < Pb−Pa F (θ ) θ=0 < 0, and F (θ )θ=1 > 0, the ESS is θ = 0.

gk−Pa , then F(γ ) = 0, indicating each γ is stable. From Eq. (3.21), if θ = Pb−Pa−Sc−Ek gk−Pa If θ = Pb−Pa−Sc−Ek , then F(γ ) = 0, indicating that there are two points of stability at γ = 0 and γ = 1. The ESS asks F’(γ ) < 0, and the derivative for F(γ ) is expressed as follows:

F (γ ) =

dF(γ ) = (1 − 2γ )[θ (Sc + Ek + Pa − Pb) + gk − Pa] dγ

(3.23)

This study only focuses on the conditions where the consumers are intended to purchase carbon labelled products, i.e. gk ≥ Pa. 1.

If gk = Pa:

92

3 Interaction Among Stakeholders Involved in Carbon Labeling Scheme

➀ ➁ 2.

  When Sc + Ek + Pa − Pb < 0, F  (γ )γ =0 < 0, and F  (γ )γ =1 > 0, then the ESS is γ = 0;   When Sc + Ek + Pa − Pb > 0, F  (γ )γ =0 > 0, and F  (γ )γ =1 < 0, then the ESS is γ = 1.

If gk > Pa: ➀ ➁ ➂

gk−Pa When Pb − Pa − Sc − Ek < 0 and Pb−Pa−Sc−Ek < 0, with the constant gk−Pa θ > Pb−Pa−Sc−Ek , the ESS is γ = 1; gk−Pa When 0 < Pb − Pa − Sc − Ek < gk − Pa, and Pb−Pa−Sc−Ek > 1, with gk−Pa constant θ < Pb−Pa−Sc−Ek , the ESS is γ = 0; gk−Pa , nWhen 0 < gk − Pa < Pb − Pa − Sc − Ek, and when θ > Pb−Pa−Sc−Ek       F (γ ) γ =0 > 0, and F (γ ) γ =1 < 0, then the ESS is γ = 1; When   gk−Pa θ < Pb−Pa−Sc−Ek , and when F  (γ )γ =0 < 0, and F  (γ )γ =1 > 0, then the ESS is γ = 0.

From the above analysis, it is clear that different ESS exits for each player under variously initial conditions. By taking actual situation into account, the study emphasizes the stability analysis of the mixed strategic actions given in the conditions where gk−Pa < 1 and 0 < Cb−Ca−Sm < 1. 0 < Pb−Pa−Sc−Ek Pb−Pa Local stability of a Jacobian matrix is employed to judge the stability of the Nash equilibrium. Nash equilibrium is a stable state, given that a game makes the unique predication from the possible strategic actions that each player may choose, and the predicted strategy should be a best response compared with the strategies chosen by all other players (Nash 1951). Any feasible payoff profile that strictly dominates the minimax profile is Nash equilibrium in an infinitely repeated game, which indicates that any strategic game has at least one pair of Nash Equilibrium (Li et al. 2016). When the equilibrium point satisfies the matrix determinant det(J) > 0, and the matrix trajectory tr(J) < 0, it can be deemed as the local stable fixed point, corresponding to the ESS, which expresses excellent disturbance resistance (Friedman 1991). In such context, evolutionary stable states in the evolutionary game are a subset of Nash equilibria (Cressman   2003). F(θ ) Let X = = f (X ) = 0, the equilibria of the game are: F(γ )          ∗  0 0 1 1 θ = X2 = X3 = X4 = X5 = X1 = γ∗ 0 1 0 1 . gk−Pa < 1, 0 < Cb−Ca−Sm < 1. Herein, 0 < Pb−Pa−Sc−Ek Pb−Pa The Jacobian matrix can be expressed as:

gk−Pa Pb−Pa−Sc−Ek Cb−Ca−Sm Pb−Pa



3.3 A Game Between Consumers and Enterprises

93

Table 3.10 Stability analysis for the equilibrium points Equilibrium point

det(J )

tr(J )

Result

(0, 0)

−

Undetermined

Saddle point

(0, 1)

−

Undetermined

Saddle point

(1, 0)

−

Undetermined

Saddle point

(1, 1) 

−

Undetermined

Saddle point

+

0

Center point

gk−Pa Cb−Ca−Sm Pb−Pa−Sc−Ek , Pb−Pa



  ∂ F(θ) ∂ F(θ) ∂ f (X ) ∂θ ∂γ = ∂ F(γ ) ∂ F(γ ) J(X ) = ∂X ∂θ ∂γ   ⎡ ⎤ γ (Pb − Pa)+ . − 2θ θ − θ − Pa) (1 ) (1 )(Pb ⎢ ⎥ Sm + Ca − Cb ⎢ ⎥   ⎣ θ (Sc + Ek − Pb + Pa) ⎦ γ (1 − γ )(Sc + Ek + Pa − Pb) (1 − 2γ ) +gk − Pa The stability of equilibrium points are given in Table 3.10. In this game, there are four saddle points: X 1 , X 2 , X 3 , and X 4 , and one center point, X 5 .

3.3.3 Evolutionary Stability Analysis by Simulation Let the enterprises’ Nash equilibrium value θ ∗ as the initial value of the probability related to the consumers’ strategic action, where the consumers’ initial value of strategic action is set as γ = 0.3, γ = 0.7, respectively. Figure 3.3 shows that the

Fig. 3.3 Evolutionary process where the consumers choose to buy

94

3 Interaction Among Stakeholders Involved in Carbon Labeling Scheme

The probability that enterprises choose to apply carbon labelling

1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

The probability that consumers choose to buy

Fig. 3.4 Evolutionary process for mixed strategic actions

probability that the consumers choose to purchase carbon labelled products fluctuates throughout the simulation period, which cannot be approached to the center point X 5 . Furthermore, for different initial values γ , e.g., γ = 0.3 (curve 1) and γ = 0.7 (curve 2), the degree of fluctuation varies. As the rounds of game increases, it is clear that the degree of fluctuation shows an increasing trend. As carbon labelled products are newly emerged products in the market, both consumers and enterprises may have a different degree of risk perception (Zhao et al. 2016). For instance, consumers may behave in a puzzled way to the carbon labelled product, because the current label information transmitted may not provide a sufficiently meaningful low carbon message (Zhao et al. 2012a). As the enterprises are voluntary to implement a carbon labelling scheme, any change may give rise to an economic risk (Zhao et al. 2017). In such case, both of the players will have a low willingness to accept carbon labelled products in the preliminary stage of the game. When both players have a probability value of 0.3 (i.e., X 5 = 0.3) as the initial value for their strategic choice, the evolutionary process of the game is shown in Fig. 3.4. It is obvious that the evolutionary process is in a periodically circular shape that begins at the initial value, reflecting that the strategic actions regarding the enterprises and consumers are bound by periodic changes, and there is no evolutionary stable equilibrium. For any external interference, it may have an impact on either of the player’s decision making.

3.3.4 Impact of Governmental Subsidy Incentives From the above analysis, it is difficult to achieve a stable equilibrium for both players, by dependence on their voluntary participations in developing the market of carbon

3.3 A Game Between Consumers and Enterprises

95

labelled products. In such premise, governmental subsidy is introduced, i.e., implementation of a dynamic subsidy for the enterprises that applies carbon label, to encourage their low-carbon production. Prior studies have indicated that incentive mechanism is efficient to promote carbon labelling practice, especially capital subsidies, by which it is clearly beneficial to foster the development and adoption of green technology to reduce carbon emissions (Mahlia et al. 2013; Dayaratne and Gunawardana 2015). In the initial stage of the incentive, government aims to provide higher subsidies to reduce the additional cost to the enterprises for conducting carbon labelling attempt. As the number of enterprises that are certified by carbon label gradually increases, government considers reducing the intensity of subsidies. Given this background, this study assumes that governmental subsidies provided to the enterprises and the enterprises that implement the labelling strategy are inversely proportional. Two categories of direct subsidies are defined, namely static and dynamic subsidies to measure their influences on enterprises’ attempt to carbon footprinting. For the former, it is set as a constant throughout the simulation period, i.e., a fixed subsidy per carbon labelled product, since a flat rate subsidy policy has been widely implemented in emerging economies (Tang et al. 2015). With regard to dynamic subsidies, it is considered as a flexible subsidy, which means that government provides a higher subsidy in the initial period of simulation, and a lower subsidy in the following period (Wang et al. 2014). In this premise, Sm is replaced by S(θ ) = (1 − θ )α in the game model, wherein a represents the upper limit of the subsidy, and 0 < α < Cb − Ca. Similar to the analysis in Sect. 4.3, the stability analysis of equilibrium points is given in Table 3.11. There are still four saddle points for the two players, i.e., X 1 , X 2 , X 3 , and X 4 . However, X 5 has been transformed into a fixed point of asymptotic stability. Figure 3.5 reflects the evolutionary process of the enterprises’ strategic action in compliance with the dynamic subsidies (curve 1), and the static subsidies (curve 2). With regard to the static subsidies, the probability that enterprises choose to apply carbon label fluctuate as the rounds of game increases. Conversely, the probability fluctuates in the initial period and gradually becomes stable when being impose by the dynamic subsidies, ultimately converges in Nash equilibrium.

Table 3.11 Stability analysis for the equilibrium points Equilibrium points

det(J )

tr(J )

Result

(0,0)

−

Undetermined

Saddle point

(0,1)

−

Undetermined

Saddle point

(1,0)

−

Undetermined

Saddle point

(1,1)

−

Undetermined

Saddle point

+

−

ESS



Cb−Ca−S gk−Pa Pb−Pa−Sc−Ek , Pb−Pa

 θ∗



96

3 Interaction Among Stakeholders Involved in Carbon Labeling Scheme

Fig. 3.5 Evolutionary process where the enterprises choose to apply carbon label

0.65 0.6 0.55 0.5

carbon labelling

The probability that enterprises choose to apply

Figure 3.6 shows the evolutionary process that both players have a probability 0.3 as the initial value of their strategic actions, under the dynamic subsidies. It is obvious that the game evolves into a spiral pattern, as the rounds of game increases, gradually converges to Nash equilibrium. This phenomenon indicates that the game has asymptotic stability since implementation of dynamic subsidies.

0.45 0.4 0.35 0.3 0.25 0.2 0.15 0.25

0.35

0.45

0.55

0.65

The probability that consumers choose to buy Fig. 3.6 Evolutionary process for mixed strategic actions under governmental subsidy

0.75

3.3 A Game Between Consumers and Enterprises

97

3.3.5 Summary Through the stability analysis of the ESS, it is obvious that the strategic actions of both players are bound by periodic changes, which reflects that the market for carbon-labeled products is vulnerable to the external factors, e.g. incentive or punitive policies. Since conducting the dynamic subsidies, the strategic actions of both players are approached to the ESS, which shows a positive motivating impact. The game theoretical analysis shows that under normal market conditions, the two players with bounded rationality are hard to find an equilibrium state. Possible reasons are given as follows: (1) consumers have a comparatively low perception of carbonlabeled products, and especially, a certain premium exists, which may give rise to uncertainty in purchasing (Upham et al. 2011); (2) the enterprises cannot precisely predict the market demand for carbon-labeled products, due to the uncertainty of consumers’ decision making. In such context, it is further suggested that the consumers with better education and positive environmental awareness, are encouraged as the precursors to be involved in buying the carbon labelled products, then to interact with other consumers by providing feedback of their purchasing experiences, in order to promote the low carbon consumption. With increasing awareness of green consumerism, enterprises may seek business opportunities from the sales of “green” products, e.g., carbonlabeled products, to undertake additional social responsibilities, ultimately to transform themselves into green businesses (Zhao et al. 2013). For example, a number of enterprises not only focus on improvement of their internal environmental performance, but also seek opportunities on low-carbon development of external supply chains (Thongplew et al. 2014; Stolka 2016). In addition, government acts as the social leader, plays an important role to lead green corporate transformation, and advocates social sustainable development (Michelsen and De Boer 2009; Zhu et al. 2013). This study takes the case of dynamic subsidies to identify that they can efficiently help both players achieve equilibrium in a short term, which demonstrates positively regulatory effect via government incentive policies on the market for carbon-labeled products.

3.4 A Game Among Enterprises This section proposes an evolutionary game theoretical approach that models the likely behavior of enterprises in response to a number of governmental policy instruments such as financial subsidies and taxation related to the implementation of a carbon labeling scheme. System dynamics is applied to simulate the created game model followed by two major scenarios, in which enterprises’ actions are determined by the individual and combined policy interventions. The application of game theory is expected to help enterprises take positive actions toward carbon emissions

98

3 Interaction Among Stakeholders Involved in Carbon Labeling Scheme

reduction by implementing a carbon labeling scheme, thereby providing insight into the design of sustainability policies that promote low-carbon development.

3.4.1 Game Theory Based SD Model Organizational behavior is driven by internal and external factors such as competition, policy instruments, and consumer demand (Tian et al. 2014), as shown in Fig. 3.7. The two players involved in the game (i.e., any enterprise in the market) are considered to have bounded rationality. Each enterprise proposed in this study has two strategic options. The first option is to implement the carbon reduction labeling scheme (hereinafter referred to as “Implement”), including the application of carbon labeling certification, low-carbon technologies, and so on (Shuai et al. 2014). The second is not to implement the carbon labeling scheme (hereinafter referred to as “N-implement”). Table 3.12 shows the payoff matrix of the two players.

Fig. 3.7 Driving factors of enterprises’ behavior

Table 3.12 Payoff matrix of the enterprises Enterprise 1

Enterprise 2 Implement

Implement

N-implement  k    ((P11 −C11 +St )× +D(St ) + 1 ; ((P12 −C12 +St )× D(ξ12 ) +D(St ) + k1 ;    ((P21 −C21 +St )× D(ξ21 ) +D(St ) + k2 (P22 − C22 ) × D(ξ22 ) 

D(ξ11 )

N-implement (P13 − C13 ) × D(ξ13 ); (P14 − C14 ) × D(ξ14 );   3 k ((P23 − C23 + St ) × D(ξ2 + D(St )]+2 (P24 − C24 ) × D(ξ24 )

3.4 A Game Among Enterprises

99

In Table 3.12i represents the enterprises that are competitive in the market, i = 1, 2; j represents the number of enterprises that have implemented the carbon labeling scheme based on the different strategic choices selected, j = 1, 2, 3, 4; t represents the different types of subsidies, t = 1, 2; k represents the different types of tax rates, k = 1, 2, 3, 4; ξ i j represents the market share of the ith enterprise based on the jth market scenario; Pij represents the unit price of the product provided by the ith enterprise based on the jth market scenario; C ij represents the unit production cost of the ith enterprise based on the jth market scenario; S t represents the t-th type of subsidy (i.e., whether it subsidizes the enterprise or consumer); Pi k represents the tax reduction for the ith enterprise from the kth type of rate; D(ξ i j ) represents the product demand of the ith enterprise based on the jth market scenario; and D(S t ) represents the increasing product demand from the t-th type of subsidy for consumers. Let x i represent the proportion of enterprises that select the strategy “Implement” among all enterprises, while the proportion of enterprises that select “N-implement” is 1 − x i . The expected payoff when enterprises choose “Implement” is set as U Ei , ¯ and the expected payoff of the average expected payoff of enterprises is set as U, enterprises choosing “N-implement” is U EN. The replicator dynamic equations of the “Implement” and “N-Implement” strategies are given as follows:   dxi ¯ = xi (1 − xi )(UEi − UEN ) = xi UEi − U dt

(3.24)

  d(1 − xi ) ¯ = (1 − xi )xi (UEN − UEi ) = (1 − xi ) UEN − U dt

(3.25)

According to the different strategic choices that result from the interactions among the conflicted players, multiple scenarios are established. SD is used to construct a causal loop system for the scenario analysis (Yunna et al. 2015). The SD model is composed of two subsystems, namely the enterprises’ subsystem and consumers’ subsystem, as shown in Fig. 3.8. Consumers’ preferences toward carbonlabeled products may have a strong influence on the enterprises’ income. Further, governmental policy is set as a moderator variable in both subsystems. Enterprises respond to the implementation of a carbon labeling scheme by varying their quantities, as determined by the replicator dynamic equation in Eq. (3.24). In reality, because enterprises may need some time to carry out a new strategic action (Zhu and Sarkis 2006), a delay function is introduced into the proposed model. In line with Dowlatshahi (2005), the delay function is set to three months in this study. In addition, the selected strategy is generally affected by enterprises’ capabilities, resources, and market forces (Tian et al. 2014). Thus, we use γ to represent the success rate of the implementation of the carbon labeling scheme. The maximum value is 1 and the mean is 0.7, since Koufteros et al. (2005) have indicated that the success rate of new product development is approximately 60–80%. The major equations of the subsystem are given as follows:

100

3 Interaction Among Stakeholders Involved in Carbon Labeling Scheme

Fig. 3.8 SD model based on an evolutionary game

EEI = INTEG (RC, initial value)

(3.26)

RC = DBC × TNE × γ

(3.27)

DBC = DELAY1(BC, 3)

(3.28)

γ = RANDOM NORMAL(0, 1, 0.7, 0.1, 0)

(3.29)

BC =

  dxi ¯ = xi (1 − xi )(UEi − UEN ) = xi UEi − U dt

(3.30)

UEi = [(GP − GC + ES) × GNU/EEI] + RT

(3.31)

UEN = (OP − OC) × (TNC − GNU)/EEN

(3.32)

3.4 A Game Among Enterprises

101

where EEI represents the number of enterprises that have implemented the carbon labeling scheme, RC the variation in enterprises that have implemented the scheme, DBC the variation after the implementation period, TNE the total number of enterprises, γ the success rate of the scheme’s implementation, BC the initial variation in enterprises that have implemented the scheme; GP the carbon-labeled product price; GC the carbon-labeled product cost; ES the direct subsidy for enterprises that have implemented the carbon labeling scheme; GNU sales of the carbon-labeled product; RT the tax rate; OP the price of the non-carbon-labeled product; OC the cost of the non-carbon-labeled product; TNC the total number of investigated consumers; and EEN the number of enterprises that have not implemented the carbon labeling scheme. The customers’ subsystem mainly contains demand for carbon-labeled products, which is shown in Fig. 3.8. Stock is controlled by the rate of the variance of consumers, which is defined by the Lyapunov function (Kelly et al. 1998), as shown in Eq. (3.33): BR =

d(G N U ) = β × (P R + C S − G P × G N U ) dt

(3.33)

where BR represents the variation in consumers; CS the direct subsidy for consumers to purchase the carbon-labeled product; and GP the carbon-labeled product price. Here, GNU is measured as the sales of the carbon-labeled product, given as follows: G N U = INTEG (B R, initial value)

(3.34)

PR is the price premium, which can be deemed as a consumer’s willingness to pay for the carbon-labeled product, given as follows: P R = (1 + θ ) × O P

(3.35)

where θ represents the coefficient of environmental preferences and OP the noncarbon-labeled product price. Whether the carbon-labeled product is welcomed by consumers may be based on their environmental preferences. Zhao and Zhong (2015) indicated that consumers with high environmental preferences are willing to buy green products.

3.4.2 An Numerical Case Air conditioner enterprises in China are herein used as an illustrative case example to demonstrate the application of the proposed evolutionary game model. In China, air conditioners are indispensable for maintaining a comfortable and healthy interior environment in residential buildings. As a result, their use accounts for a major share

102

3 Interaction Among Stakeholders Involved in Carbon Labeling Scheme

of energy consumption in such buildings (Chua et al. 2013). Indeed, Kyle et al. (2010) identified a strong link between climate change and heating and cooling demand. The rapid development in the country’s building industry over the past few decades (Chen et al. 2015) has led to an increase in the sales of air conditioners (Fig. 3.9). However, to mitigate the impact of climate change, a series of countermeasures have been proposed to influence China’s air conditioning industry. For example, GREE Corporation invests billions of Chinese yuan annually to promote the application of fluoride-free technology when designing air conditioners (China HVACR 2015). Moreover, China’s air conditioner enterprises can be certified to verify their energy efficiency (Lin and Rosenquist 2008). Labeling the energy efficiency of air conditioners, which is a typical type of eco labeling scheme, is divided into three grades (i.e., grades 1–3). The first grade represents the lowest energy consumption (i.e., a refrigerating coefficient of less than 3.4; Standardization Administration of the People’s Republic of China 2010). Similarly, we use the methodology proposed by Zhao et al. (2012a) to divide carbon labeling into three reduction grades, namely high, medium, and low, to construct a carbon reduction labeling scheme, as shown in Fig. 3.10. The data on the input parameters of the SD model are derived from the China’s Household Electrical Appliances Association, National Bureau of Statistics of the People’s Republic of China, and similar studies (Table 3.13). The number of the enterprises is set to 32, and air conditioner sales were 43.9 million units in 2014 (National Bureau of Statistics of the People’s Republic of China 2015). Since China has not yet launched an official carbon reduction labeling scheme, the pilot enterprises using carbon labeling are considered to be using the first grade of the energy efficiency label, which accounted for 9.7% of market share in 2014 (ZOL 2015). Thus, we define

180

Sales of air conditioners (Million sets)

160 140 120 100 80 60 40 20 0 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Time (Year)

Fig. 3.9 Sales of air conditioners

3.4 A Game Among Enterprises

103

Fig. 3.10 The proposed carbon reduction labeling scheme

Table 3.13 Data on the input parameters Level variables

Constant

Parameter

Data

Unit

EEI

3

/

EEN

29

/

GNU

11.5

Million set

OC

5000

RMB

OP

5600

RMB

GC

5900

RMB

GP

6500

RMB

θ

0.1

/

TNE

32

/

TNC

43.9

Million set

the number of pilot enterprises is three. Similarly, the price of a carbon-labeled air conditioner is 16% higher than that of an ordinary one (ZOL 2015). The subsidy standard is based on the Ministry of Industry and Information Technology of the People’s Republic of China (2012), indicating that enterprises will receive a subsidy ranging from 300 to 400 RMB for each product, according to their corresponding energy efficiency grade. Lin and Jiang (2011) indicated that subsidies to different stakeholders may result in varying degrees in low-carbon consumption and production. In the same vein, consumers in this study are assumed to receive a subsidy when purchasing a carbon-labeled product. Consistent with the subsidy standard for enterprises, consumers’ subsidies are given in three hierarchies: 300 RMB, 350 RMB, and 400 RMB per unit. According to the Enterprise Income Tax Law of the People’s Republic of China, the general tax rate on the enterprise income is 25%. Preferential taxation policy aims to waive a certain percentage of the full corporate income tax rate (Huang 2006). Furthermore, the central government implements preferential tax policies for SMEs and high-tech enterprises as the rate as 15% and 20%, respectively (The Central

104

3 Interaction Among Stakeholders Involved in Carbon Labeling Scheme

People’s Government of the People’s Republic of China 2007). However, no specific preferential tax policy drives carbon labeling schemes in China. Elschner et al. (2011) and Ng et al. (2012) identified that tax reductions may serve as an incentive for green development. Hence, we define four categories of preferential taxation rate, namely a fixed tax rate, stepped tax rate, linear decreasing tax rate, and inverted triangle tax rate. The fixed tax rate is set as a constant of 20% throughout the simulation period. The stepped tax rate begins with 25% until the second stage of the simulation and then decreases to 20% between the second and the fourth stage, remaining at 15% until the end of the simulation. The linear decreasing tax rate is where the taxation decreases linearly from 25 to 15% from the start to the sixth stage and then remains at 15% until the end of the simulation. The inverted triangle tax rate decreases linearly from 25 to 15% from the start to the third simulation stage and then increases to 20% until the sixth simulation stage, remaining at 20% for the rest time of the simulation. Table 3.14 presents the diagrams of the four predefined tax rates and their corresponding functions.

3.4.3 Simulation Results The software package Vensim PLE for Windows Version 6.3 is used for the SD model simulation. Two scenarios were built to investigate how enterprises responded to different governmental policies when implementing the carbon labeling scheme. Scenario 1 assessed enterprises’ responses to different subsidy and preferential taxation (as reflected by the variation in the number of enterprises), where the policy instruments are employed separately. Scenario 2 assessed enterprises’ responses to a combined policy instrument (again, as reflected by the variation in the number of enterprises). In Scenario 1, the number of enterprises remains at a stable level (around eight), although there is a slight increase in the initial stage (Fig. 3.11). This variation trend line is set as the benchmark to enable comparison with Scenario 2. The responses of enterprises are determined by the change whether adopted the carbon labeling scheme. The initial number regarding the adopting enterprises have increased because firms want to benefit from the development of low-carbon products in the sense of improving the products’ green value, enhancing their product competiveness, and raising their corporate social image (Chen 2008). However, market demand is a critical factor in business operations (Lin et al. 2013). Consumer’s preference is a key driver of market demand, although the perceived value is diverse (Zhou et al. 2009). As the market for the low-carbon products gradually becomes saturated, consumption is expected to be decreased, which may result in a temporal fluctuation in the number of enterprises that implement the carbon labeling scheme (Janssen and Jager 2002). Further, not all enterprises are willing to incorporate green technologies into their product innovation because of unintended market risk, additional costs, and so on (Lin et al. 2013; Zhao et al. 2015). This fact may explain why

20%

WITHLOOKUP (time, ([(0,0)–(50,0.5)],(0,0.25),(2,0.25),(2,0.2),(4,0.2),(4,0.15),(6,0.15),(50,0.15)))

WITHLOOKUP (time, ([(0,0)–(50,0.5)],(0,0.25),(6,0.15),(50,0.15)))

WITHLOOKUP (time, ([(0,0)–(50,0.5)],(0,0.25),(3,0.15),(6,0.2),(50,0.2)))

Stepped tax rate

Linear decreasing tax rate

Inverted triangle tax rate

Function

Preferential tax rate

Diagram

Category

Fixed tax rate

Policy variable

Table 3.14 Determination of preferential taxation policy

3.4 A Game Among Enterprises 105

106

3 Interaction Among Stakeholders Involved in Carbon Labeling Scheme

Fig. 3.11 Responses of the enterprises to the carbon labeling scheme

the number of enterprises that adopt the carbon labeling scheme ultimately remains around eight. Figure 3.12 shows that the direct subsidy affects enterprises’ decision to implement the carbon labeling scheme. As the subsidy rises, the number of enterprises increases compared with the benchmark. When the same subsidies are given to consumers, the equilibrium values are 8, 9, and 10, respectively. Hence, a direct subsidy to the enterprises is much better than that given to the consumers, in line with the findings of Tian et al. (2014). One possible reason is that enterprises are more sensitive to incentive policies compared with consumers (Diamond 2009). For instance, enterprises have paid great attentions to the impact of carbon allowances (Zhang et al. 2015). Figure 3.13 shows that the enterprises’ responses to the four preferential taxation policies. Compared with the other three categories of preferential taxation, a fixed tax rate fosters rapid growth in the enterprises that adopts the carbon labeling scheme in the initial stage. However, the corresponding equilibrium number is the least. The stepped tax and linear decreasing tax rates give rise to the same equilibrium, although the latter encourages the enterprises to implement the scheme more effciently. In this context, the performance of the linear decreasing tax rate seems better than that of the stepped one. The dynamic equilibrium imposed by the inverted triangle tax rate approaches that of the fixed tax rate when the preferential tax rate is 20%. However, it is below the actions imposed by the stepped and linear decreasing tax rates. The linear decreasing tax rate is thus suggested as the optimal form of taxation policy because of its predictability, which may enforce the enterprises to participate in carbon labeling more aggressively.

3.4 A Game Among Enterprises

107

a. Subsidy for enterprises that have implemented the carbon reduction labeling schemeu

b. Subsidy for consumers to buy carbon-1abeled productse Fig. 3.12 Enterprises’ responses to the direct subsidy

In summary, Scenario 1 found that the linear decreasing tax rate and enterprise subsidy of 400 RMB are the optimum policy making choices. Moreover, the enterprise subsidy resulted in a greater promotion in the enterprises implementing the carbon labeling scheme in the early stage of the simulation, as shown in Fig. 3.14. However, the equilibrium number is smaller, which reflects that the linear decreasing

108

3 Interaction Among Stakeholders Involved in Carbon Labeling Scheme

Fig. 3.13 Enterprises’ responses to the four predefined preferential tax rates

Fig. 3.14 Enterprises’ responses to the two optimal incentive policies

tax rate may have long-term efficacy. This result verifies the findings of Ockwell et al. (2008), who demonstrated that preferential taxation could be more effective than a subsidy for the development of low-carbon products.

3.4 A Game Among Enterprises

109

In Scenario 2, the two incentive policies are combined to investigate the responses of the enterprises. Figure 3.15 shows that all the three policy combinations have a similar trend in terms of the number of enterprises that implement the carbon labeling scheme. As the subsidy increases, the fixed tax rate triggers the largest growth in the enterprises to implement the scheme, followed by the linear decreasing tax and stepped tax rates. The inverted triangle tax and fixed tax rates trigger nearly the same equilibrium at a 20% rate, which has the weakest influence on the changes in the enterprises. By comparing Scenarios 1 and 2, we can see that the equilibrium values are increased in the latter Scenario, as shown from Figs. 3.12, 3.13, 3.14 and 3.15. Hence, the combined policy has a greater influence on the implementation of the carbon labeling scheme. The optimum combination is a direct subsidy of 400 RMB for each air conditioner with a linear decreasing tax rate, as shown in Fig. 3.16. This combination of policies gives rise to not only a faster growth in enterprises implementing the scheme, but also a more stable market share. In particular, the equilibrium value is larger than that when choosing any individual policy.

3.4.4 Sensitivity Analysis A sensitivity analysis is carried out to investigate whether the results vary once the related parameters changing and to assess which variable has the greatest impact on the results (Blumberga et al. 2015; He and Zhang 2015). Four variables are selected for the sensitivity analysis: the coefficient of environmental preferences, the coefficient of rate control, the initial number of the enterprises that implement the carbon labeling scheme, and the total number of the consumers. The sensitivity is calculated by using the slope between −10% and 10% of the variance of the four variables to investigate the changes in the number of the enterprises. The results show that if the related parameters change from −10% to − 10%, the change in the number of the enterprises that adopts the carbon labeling scheme is within ±1%, as shown in Fig. 3.17. Since this sensitivity is in a reasonable range, the model is regarded as robust for the simulation.

3.4.5 Summary Game theoretical analysis allows the identification of alternative business strategies to help enterprises reduce their lifecycle-based carbon emissions without affecting their commercial sustainability. In particular, the simulation results show that a direct subsidy given to the enterprises is better than that given to the consumers.

110

3 Interaction Among Stakeholders Involved in Carbon Labeling Scheme

a.

Subsidy for enterprises as 300 Yuan per unit

b. Subsidy for enterprises as 350 Yuan per unit

c. Subsidy for enterprises as 400 Yuan per unit

Fig. 3.15 Enterprises’ responses to the combined policies

3.4 A Game Among Enterprises

111

Fig. 3.16 Optimal combination of the policy intervention

Fig. 3.17 Sensitivity analysis of the model

A combination of incentives (i.e. a direct subsidy and preferential taxation) is more efficient to drive the implementation of the carbon labeling scheme. When launching such a scheme, government incentives such as subsidies and preferential tax rates may play a key role in stimulating enterprises to pursue carbon labeling (Geng and Doberstein 2008; Zhao et al. 2013). Nevertheless, a continuous incentive mechanism may result in a financial burden for the government.

112

3 Interaction Among Stakeholders Involved in Carbon Labeling Scheme

Olson (2013) verified that governmental incentives are insufficient to overcome weak market demand. For long-term success in the implementation of the carbon labeling scheme, governmental incentives must be in an appropriate intensity to drive enterprises towards low-carbon production.

3.5 A Game Between Enterprises and Government The implementation of a carbon labeling scheme involves the coordination of various stakeholders such as governments, enterprises, and consumers, which may take the form of a mandatory or a voluntary system (Tan et al. 2014). This section uses system dynamics (SD) to model enterprises’ response to government policy making (e.g. combined subsidies and penalties combined) for the implementation of a carbon labelling scheme. The synergy impact of subsidies and penalties on enterprises’ economic return is investigated. Two optional technological plans for carbon emissions reduction for the enterprises are defined to determine which plan is effective in enhancing both economic and environmental performance under the policy making. An optimal combination of policy design is suggested to drive carbon emissions reduction and to promote low-carbon transition. The application of SD provides insight into the modelling of policies influencing enterprises’ responses in implementing a carbon label scheme to reduce their life-cycle based carbon emissions.

3.5.1 SD Model Construction of a SD model for policy simulation requires the following steps (Wu et al. 2011; Yunna et al. 2015): identifying the problem situation, sketching a system structure to understand the factors within the system and their inner relationships, establishing casual loops to describe the logical structure of the system, building equations among the factors to generate quantitative relationships, using computer-based software to simulate, debug, and examine the established model, and investigating the possible impact of a change regarding the control variables on the model output. This study will investigate two combined policy instruments— subsidies and economic sanctions—informing the enterprises’ technological choice of carbon emissions reduction, as shown in Fig. 3.18. Here, the enterprises are considered an entity, which has two optional technologies for the carbon emissions reduction defined as the low-cost plan (i = 1) and the high-cost plan (i = 2), respectively. The detailed instructions for these technological plans are shown in Table 3.15. Their division is according to the results of

3.5 A Game Between Enterprises and Government

113

Fig. 3.18 The proposed SD model in implementation of carbon reduction labelling scheme

Table 3.15 Division of the technological plan Technological plan

Measures

High-cost plan

Application of energy efficient technology Installation of energy efficient equipment Increased use of renewable energy Increased investment in environmental prevention

Low-cost plan

Improvement in employees’ environmental consciousness Improvement in working conditions related to health safety Reinforcement of low-carbon development training Increase in the recyclability of products and their associated components

Zeng et al. (2010) and Yusup et al. (2015), who suggested classifying the technological plans into two categories in terms of their principles of cleaner production, namely, high and low-cost plans based on their financial budgets. The low-cost plan requires little capital investment to achieve the environmental improvements, such as establishing the environmental management system and raising employees’ environmental conscientiousness, whereas the high-cost plan demands financial investments that involve redesign of production processes such as the installation of energy efficient facilities and use of clean energy (Zeng et al. 2010). In this context, the high-cost plan is implied to have greater potentials in carbon emissions reduction. This study has deliberately applied the marginal cost of carbon emissions reduction to discriminate these two plans, that is, $10 per tonne for the low-cost plan and $15 to $18 per tonne for the high-cost plan (US Environmental Protection Agency 2010; Wang and Song 2010). Either technological plan selected cannot be adjusted within a specific period, for example, from the year 2005 to 2020, which is the pre-defined period for the implementation in this study.

114

3 Interaction Among Stakeholders Involved in Carbon Labeling Scheme

The system structure emphasizes the internal causal relationship between policy regulation and the enterprises’ response. In this study, government policy instruments mainly focus on financial subsidies and sanctions, in which the former is beneficial to foster the development and adoption of green technology to reduce carbon emissions (Badcock and Lenzen 2010; Zhao et al. 2015). The latter compels enterprises to implement low-carbon production and helps government impose necessary surveillance on enterprises’ environmental unfriendly behaviours (Tsou and Wang 2012; Zhao et al. 2013). Thus, the combination of subsidies and sanctions is an important indicator to represent the policy regulation. Because enterprises are driven by profits, enterprises’ economic returns are indicated as a decision variable to measure whether they may like to attempt carbon labelling or not. Figure 3.19 shows the causal loop diagram corresponding to the SD model. There are six causal feedback loops in the system, which are centred on the enterprise’s economic return. For instance, environmental investment follows a positive correlation; as economic return increases, the technological level rises. As indicated by Sloan (2011), and Sgobbi et al. (2016), green technological innovation

Fig. 3.19 Causal loop diagram of the SD model

3.5 A Game Between Enterprises and Government

115

has a strongly positive impact on the reduction of carbon emissions. With improvements in green technology, the carbon emissions intensity (per unit of product) gradually decreases to have a positive impact on the reduction of carbon emissions. Greater environmental investment, therefore, updates technological level for carbon emissions reduction, which has been verified by the results in Lu et al. (2010). Additionally, government drives the market by imposing incentive instruments to internalize the negative effect of greenhouse gas emissions by providing a net revenue constraint (Galinato and Yoder 2010). Thus, the greater the reduction efficiency, the more subsidy obtained, which is described as a positive feedback loop. The economic return has a positive influence on expanded production, which further increases the supply–demand ratio, indicating as a positive feedback loop. Such feedback loop has been verified by the results in Iles and Martin (2013) and Babar et al. (2016): extended reproduction mainly depended on the net profit of an enterprise, which could vary the supply–demand relationship. With consumer demand increasing, product sales are promoted to create a positive feedback on the economic return (Dijk and Yarime 2010; Harrison et al. 2014). However, there is a negative relationship between the unit price of a product and the quantity demanded, which affects the fluctuation of the supply–demand ratio (Kocabiyikoglu and Popescu 2011; Avinadav et al. 2013; Babar et al. 2016) (Fig. 3.19).

3.5.2 An Illustrative Case China’s paper-making enterprises represent a typical case example for the SD simulation. It is assumed that the paper-making enterprises in the model are entirely rational to maximize their economic returns in implementation of the carbon labelling scheme. Additionally, the enterprises are hypothesized to accept heavy economic sanctions instead of closure, even with large quantities of carbon emissions. Establishing a SD model follows a series of procedures, one of which is to build causal feedback loops based on the inner factors to explain their relationships, as shown in Fig. 3.19 (Feng et al. 2008; Wu et al. 2011; Yunna et al. 2015). These factors are divided into a number of variables, that is, level, rate, and auxiliary variables, which are based on the decomposition of the system objective (Wu et al. 2011; Tian et al. 2014). Similarly, all the inner factors in our model have been defined as corresponding variables, which are shown in Table 3.16. In a SD model, the level variable represents the accumulation of any variable at a specific time. The level increases or decreases at a rate that depends on the inflow or outflow (Suryani et al. 2010). For instance, the economic return in our study is considered as a level variable to record its accumulated transformation during a period of simulation time. Economic return is affected by the economic income and economic expenditure in accordance with the rate variable. The auxiliary variable is intended to link level and rate variables for intermediate calculations (Ahmad et al. 2015). The supply–demand ratio, economic

116

3 Interaction Among Stakeholders Involved in Carbon Labeling Scheme

Table 3.16 Key variables of the SD model Variable

Type

Definition

Units

Economic return (ER)

Level

The paper enterprises’ total economic return

Million yuan

Price (P)

Level

Product sales price

Yuan

Per unit cost of production (UC)

Level

The embedded cost of production

Yuan

Total population (TP)

Level

Population gross

Million

Total capacity (TC)

Level

Capacity of production

Million tonnes

Carbon emissions (CE)

Level

The enterprises’ total carbon emissions annually

Million tonnes

Carbon emissions per unit of product (UCE)

Level

The carbon emissions per product

Tonnes

Annual growth of population (PG)

Rate

Average growth of population

Million

Newly increased capacity (NCR)

Rate

Increased capacity annually

Million tonnes

Increment of cost (CI)

Rate

Annual increment of cost Yuan

Economic income (I)

Rate

General income

Million yuan

Economic expenditure (E)

Rate

Total expenditure

Million yuan

Variance of carbon emissions per unit of product (UCV)

Rate

The carbon emissions reduction per unit of product

Tonnes

Annual rate of population increase (PR)

Auxiliary

Average growth rate of population

/

Demand per capita (PD)

Auxiliary

Annual demand per capita

Kg per capita

Total demand (TD)

Auxiliary

Total demand annually

Million tonnes

Yield (Y)

auxiliary

Enterprises’ total production

Million tonnes

Supply–demand ratio (RA)

Auxiliary

The yield divided by the aggregate demand

/

Growth rate of capacity (GC)

Auxiliary

Annual growth rate of capacity

/

Growth rate of capacity planning (PCR)

Auxiliary

Annual growth rate of capacity planning

/

Rate of outdated capacity reduction (OCR)

Auxiliary

Annual rate of outdated capacity reduction

/

Growth rate of cost (CR)

Auxiliary

Annual growth rate of cost

/ (continued)

3.5 A Game Between Enterprises and Government

117

Table 3.16 (continued) Variable

Type

Definition

Factor of supply–demand ratio (SDF)

Auxiliary

Influence factor between / supply–demand and price

Price adjustment coefficient (PAC)

Auxiliary

Price adjustment / coefficient resulted from inflation

Economic subsidy (ESS)

Auxiliary

The subsidy for carbon emissions reduction

Million yuan

Economic sanction (ESC)

Auxiliary

The penalty for high carbon emissions

Million yuan

Environmental investment (EIV)

Auxiliary

Total investment related to environmental activities

Million yuan

Carbon emissions intensity (CEI)

Auxiliary

The carbon emissions Tonnes/million yuan divided by the economic return

Intensity of economic subsidy (SS)

Auxiliary

Subsidy per carbon emissions reduction

Yuan/tonne

The possible economic penalty for compulsive emissions reduction

/

Intensity of economic sanction Auxiliary (SI)

Units

subsidy, and investment are described as auxiliary variables to aid the quantification of economic income and expenditure. The parameters related to the SD model are external inputs, and their measurement is rooted in a number of common statistical approaches, for example, table function, arithmetic method, trend method, and regression method, as shown in Table 3.17. The table function is mainly used to represent a set of data that depend upon time variation to reflect a nonlinear relationship between variables versus time (Jin et al. 2009; Guo and Guo 2015), which is used in this study for the targeted carbon emissions intensity and environmental investment. The arithmetic method decomposes aggregating data into their constituent parts and generates an overall mean prediction (Cornillie and Fankhauser 2004; Zhang et al. 2009), by which it has been applied to the variable of the price adjustment coefficient and the factor of the supply–demand ratio. The trend method provides an overall linear trend estimate to describe the change in the investigated variable over the entire study period (Soldaat et al. 2007; Hsu 2012), for example, it applies to the measurement of the variable regarding the annual population increase rate. The regression method, which is used to explain the potential relationships between independent variables and dependent variables, has been widely applied to the quantification of product demand, such as water demand (Qi and Chang 2011) and electricity demand (Bianco et al. 2009). In the study, the paper demand is measured by the regression method. The essential input data, such as the annual economic return, carbon emission, and growth of carbon emissions are

118

3 Interaction Among Stakeholders Involved in Carbon Labeling Scheme

Table 3.17 Measurements of external input model parameters Methods

Parameter

Measurement

Arithmetic method

Price adjustment coefficient (PAC)

PAC = PI(t) /PI(t−1) where 1.03 PI(t) indicates the price index in the year of t, PI(t) the price index in the year of t−1

Factor of supply–demand ratio (SDF)

SDF = Y/TD where Y is 0.16 the yield, TD the total demand

Growth rate of cost (CR)

CR = [UC(t) − 0.12 UC(t−1) ]/UC(t−1) where UC(t) is the cost in the year of t, UC(t−1) the cost in the year of t−1

Annual rate of population increase (PR)

PR = [TP(t) − TP(t−1) ]/TP(t−1) where TP(t) is the population in the year of t, TP(t−1) the population in the year of t−1

Growth rate of capacity planning (PCR)

PCR = PC/TC where PC 0.07 is the capacity planning, TC the total capacity

Rate of outdated capacity reduction (OCR)

OCR = OC/TC where OC is the outdated capacity, TC the total capacity

Trend method

Table function method

Regression method

Numerical value

0.0048

0.025

Targeted carbon emissions TCI = CE/CR where CE intensity (TCI) tonnes per is the enterprises’ carbon Million yuan emissions, CR the enterprises’ economic return

(2005,4500) (2010,4000) (2015,3500) (2020,2000)

Environmental investment (EIV) Million yuan

Low-cost technological plan: (2005,750) (2006,50) (2020,50) High-cost technological plan: (2005,2500) (2006,100) (2020,100)

Demand per capita (PD)

PD = 5.336x-10657 R2 = 0.985

3.5 A Game Between Enterprises and Government

119

sourced from the National Bureau of Statistics, China Paper.com, and China Paper Association.

3.5.3 Simulation Results Once the SD model has been built, a validity check is required (Peterson and Eberlein 2006). The simulation results of the two important indicators, i.e. the enterprises’ economic return and the carbon emissions, are compared with their historical data for model validation. Table 3.18 shows the differences between the simulation results and the historical statistics in the past five years, for which the relative error is shown within 15% that can be considered as valid (Tang et al. 2012). Thus, the simulation results can be set as a benchmark to further investigate the possible variation in economic returns from a change resulting from policy combination. Two case scenarios are established in which the policy variables are pre-defined. The economic subsidy is indicated by Yuan per tonne of carbon emissions reduction, and the intensity of economic sanctions is expressed by the proportion of economic penalties accounted for the economic return, as shown in Table 3.19. In Scenario 1, the enterprises’ economic return is examined by the change related to the combined policy of subsidies and sanctions. The enterprises’ economic return may be lower or higher than the benchmark, which can be indicated by the relative ratio of the financial loss or increment, given as follows:       L Sk , Pj , t = [R(t) − R Sk , Pj , t ]/R(t) *100%

(3.36)

Table 3.18 A comparison of the simulation and historical value of the paper industry Parameters Economic return (Billion yuan)

Carbon emission (million tonne)

2005

2006

2007

2008

2009

Historical value

1.18

1.51

2.1

2.1

2.1

Simulation result

1.18

1.44

1.85

2.13

2.4

Relative error (%)

0

−4.6

−11.9

1.4

14.2

Historical value

87.88

92.3

97.7

107.2

113.9

Simulation result

91.1

100.09

111.01

122.60

130.09

Relative error (%)

3.4

8.3

13.7

13.8

14.3

Table 3.19 Numerical value of different policy variables Policy factor Low-cost plan

Economic subsidy

Value 1

Value 2

Value 3

Value 4

60

120

180

240

0.01

0.013

0.016

0.02

High-cost plan Intensity of economic sanction

120

3 Interaction Among Stakeholders Involved in Carbon Labeling Scheme

      E Sk , Pj , t = [R Sk , Pj , t − R(t)]/R(t) *100%

(3.37)

where L(S k , Pj , t) and E(S k , Pj , t) represent the relative ratio of financial loss and financial increment in year t when the subsidy and economic penalty are set as S k , Pj , respectively. R(t) is the benchmark value in the year t, and R(S k , Pj , t) is the enterprises’ economic return in the year t when the subsidy and economic penalty are set as S k , Pj , respectively. Specifically, S k = (6, 120, 180, 240), and Pj = (0.01, 0.013, 0.016, 0.02). The averagely relative ratio of the financial loss and increment are given as follows: n    L Sk , Pj , t /n La =

(3.38)

i=1

Ea =

n    E Sk , Pj , t /n

(3.39)

i=1

where L a and E a denote the average relative rate of financial loss and increment from the year 2005–2019, n = 15. The enterprises are exposed to financial losses in the initial simulation stage, because fewer subsidies are offered by the government. In the case of 60 Yuan being offered as a subsidy, the enterprises’ average ratio of economic loss is expected to vary from 7.5 to 14% with economic penalties increasing from 0.01 to 0.02, as shown in Fig. 3.20a. When the subsidy is raised to 120 Yuan per tonne, the enterprises are still on the verge of economic loss, as shown in Fig. 3.20b. Particularly, the maximally relative ratio of economic loss remains at 9% when the intensity of the economic sanction is set as 0.013. A possible reason is that enterprises may suffer from shortages in technological innovation funding (Jin and Zhang 2014; Zhang et al. 2016). If the subsidy cannot totally cover the additional cost, it is demotivating for mid and small-sized enterprises to take voluntary action to reduce carbon emissions. As the subsidy increases, the economic return gradually increases. When the subsidy becomes 180 Yuan per tonne, the maximally relative ratio of economic return is further decreased to 7.9%, shown in Fig. 3.20c. However, the enterprises are possibly facing economic loss if the intensity of economic sanction is high enough, that is, the intensity is set between 0.013 and 0.02. If the subsidy is increased to 240 Yuan, the enterprises’ economic return ultimately surpasses the benchmark (line 5) with a maximally relative ratio for the economic return as 11.4%, shown in Fig. 3.20d. Compared with the subsidy, the intensity of economic penalties has less impact on the enterprises’ economic return, which is shown in Fig. 3.21. A possible reason might be that the enterprises take positive actions on technology upgrading to improve their green performance to avoid heavy penalties on their environmentally unfriendly behaviour (Zhao et al. 2012b; Zhu et al. 2013). If the subsidy is above 180 Yuan, the enterprises’ economic return is above the benchmark value. The maximally relative

3.5 A Game Between Enterprises and Government

Fig. 3.20 Variance in the economic return with different economic subsidies (low-cost plan)

Fig. 3.21 Variance in the economic return with different economic sanctions (low-cost plan)

121

122

3 Interaction Among Stakeholders Involved in Carbon Labeling Scheme

Fig. 3.22 Variance in the economic return with different economic subsidies (high-cost plan)

ratio of the economic return is 22.7% when the intensity of economic sanctions is set as 0.01. Compared with the low-cost plan (Scenario 1), the enterprises may face greater economic risk in choosing the high-cost plan because additional one-off investments are required to improve product sustainability and promote the low–carbon transition. If the subsidy is not high enough, the economic return will be lower than the benchmark value. Figure 3.22a shows that the maximally relative ratio of the economic loss is approximately 14.4% (indicated by line 4, the intensity of economic sanction is set at 0.02) when the subsidy is set at 60 Yuan. When the subsidy is raised to 120 Yuan, the maximally relative ratio of the economic loss is decreased to 6.8%, shown in Fig. 3.22b. When the subsidy is increased to 180 Yuan, the enterprises still face economic loss, with the maximally relative ratio of economic loss decreasing to 2%, shown in Fig. 3.22c. Only if the subsidy is raised to 240 Yuan does the economic return of the enterprises’ rise above the benchmark value with a maximally relative ratio of the economic increment at 1.5%, shown in Fig. 3.22d. When the intensity of economic penalties and subsidies are set at 0.01, 240 Yuan, respectively, the maximally relative ratio of the economic increment is 1.3% (indicated by line 4), shown in Fig. 3.23a. However, the expected economic return decreases as the intensity of economic penalties increases. Unless the subsidy is sufficiently high, that is, 240 Yuan or more, the enterprises face a loss of varying degrees

3.5 A Game Between Enterprises and Government

123

Fig. 3.23 Variance in the economic return with different economic sanctions (high-cost plan)

with an increase in the intensity of economic penalties, as shown in Fig. 3.23b, c. Particularly, the maximally relative ratio of the economic loss is approximately 18%, which is reflected in Fig. 3.23d. It is thus suggested that the penalties should not be extremely severe while taking enterprises’ cost affordability into account (Agan et al. 2013; Suk et al. 2014). Otherwise, low-carbon transition may be demotivating.

3.5.4 Sensitivity Analysis and Discussion The sensitivity analysis observes the variance in system influence on model output (Sterman 2000). The sensitivity is calculated by the slope between −15 and 15% of the variance in the control variables, that is, the intensity of the subsidies and economic sanctions, to measure their influence on the enterprises’ economic return, shown in Fig. 3.24. The larger the absolute value of the slope, the higher the sensitivity of the variables. The x-axis indicates the variance in the control variables, and the yaxis indicates the variance in economic return. The sensitivity is shown in a reasonable range, indicating that the SD model is robust for the simulation. Both of the technological plans are vulnerable to the combined policy making. When any technological plan is selected, the carbon emissions reduction is intended to be determined, indicated by the relative ratio of carbon emissions reduction, which is measured as follows:

124

3 Interaction Among Stakeholders Involved in Carbon Labeling Scheme

Fig. 3.24 Sensitivity analysis of the SD model

Fig. 3.25 Variance in the carbon emissions with the low-cost plan and the high-cost plan

Cr (i, t) = {[C(t) − C(i, t)/C(t)]} ∗ 100

(3.40)

where C r (i,t) denotes the relative reduction ratio by using the ith technological plan in the year t, C(t) is the benchmark for the carbon emissions reduction in the year t, tonne, C(i,t) is the carbon emissions reduction by using the ith technological plan in the year t, either high-cost or low-cost plan.

3.5 A Game Between Enterprises and Government

125

Figure 3.25 shows that emissions reduction from the high-cost technological plan is apparently superior to that of the low-cost technological plan. The average relative reduction ratio is 17.8% with the low-cost plan and 24.5% with the high-cost plan. However, the plan selection is mainly depended upon the enterprises’ economic stability and compliance with policy making (Pan et al. 2015). As the traditional basis on enterprises’ operation is governed by the profit motive, there would be a tendency to neglect their higher moral responsibility to achieve green societal value for environmental protection (Su and He 2010). The simulation results indicate that the paper mill enterprises in the current stage may prefer implementing the low-cost plan. A possible reason is that the mid- and small-sized enterprises account for a large proportion, and they may not assume such a heavy one-off investment from start-up to improve their environmental performance (Liu et al. 2013; Kong et al. 2013). In addition, a flat rate subsidy policy is widely implemented in China, by which the two optional plans are undifferentiated (Tang et al. 2015). The existing incentive policy is identified as ineffective to stimulate the enterprises to implement the high-cost plan. Although the high-cost plan shows superior performance on carbon emissions reduction, the selection of the low-cost plan may be more rational for the enterprises in terms of maximizing their economic returns. For instance, Fig. 3.26 shows that the maximum ratio of the economic loss is expected to be 4.5%, which is far less than the maximum ratio of the economic loss under the high-cost plan, as 10.2% shown in Fig. 3.27. When the low-cost plan is selected for further implementation, subsidies are significant to incentivize the enterprises to maximize their economic return, as shown in Figs. 3.20 and 3.21. To compensate for possible losses, the minimum subsidy

Fig. 3.26 Variance in the economic return with policy combination (low-cost plan)

126

3 Interaction Among Stakeholders Involved in Carbon Labeling Scheme

Fig. 3.27 Variance in the economic return with policy combination (high-cost plan)

should be at least 120 Yuan. When the subsidy is increased to 240 Yuan, the maximum economic return can be realized by the enterprises. However, such a high subsidy may result in fiscal stress on the government, possibly reducing the governmental revenue (Suk et al. 2013; Craig and Allen 2014). In terms of the carbon emissions intensity, Fig. 3.28 shows that line 3 (subsidy set at 180 Yuan) is approximately overlapping with line 4 (subsidy set at 240 Yuan), which reflects similar efficiency in emissions reduction. To help enterprises improve their environmental performance as well as mitigate governmental fiscal stress, we suggest that the optimal subsidy is 180 Yuan. This implication is verified by recent economic incentives in China, which indicates that the central government is intended to increase the intensity of subsidy from 240 to 500 Yuan per tonne of standard coal reduction (China Economic Herald 2013). As one tonne of standard coal may cause 2.67 tonnes of carbon emissions, the subsidy is converted to 187 Yuan per tonne of carbon emissions reduction. If the optimal subsidy is taken as 180 Yuan as well as the intensity of economic sanction is set as 0.01, the enterprises may benefit from the minimum carbon emissions intensity, as shown in Fig. 3.29.

3.5.5 Summary Government often restricts the market by legislation or regulation, sometimes with unintended consequences (Luo et al. 2010). This study attempts to explore one such consequence which has particular relevance to carbon emissions reduction, e.g. how

3.5 A Game Between Enterprises and Government

127

(a) The intensity of economic sanction is set at 0.01

(b) The intensity of economic sanction is set at 0.013

(c) The intensity of economic sanction is set at 0.016

(d) The intensity of economic sanction is set at 0.02

Fig. 3.28 Variance in the carbon emissions intensity with different economic sanctions (low-cost plan)

Fig. 3.29 Variance in the carbon emissions intensity with the subsidy fixed as 180 Yuan per tonne of carbon emissions reduction (low-cost plan)

128

3 Interaction Among Stakeholders Involved in Carbon Labeling Scheme

governments establish a reasonable incentive and restraint mechanism to encourage enterprises to improve their environmental performance by reducing their lifecycle based carbon emissions. The presented simulation results give an optimal combination of economic subsidy and sanction, to offer insight into government policy in setting the appropriate intensity of subsidies and penalties to accelerate low-carbon development. The results provide implications for the enterprises to select the optimal strategy to improve their environmental performance in the initial stage of carbon labelling practice. With low-carbon production activities gaining increasing influence on the market, enterprises are expected to benefit from business opportunities connected with sales of carbon labelled products (Zhao et al. 2013, 2015). In this context, it is considered that the cost related to technological innovation for carbon labelling accreditation may be totally covered by the enterprises. If government further raises standard of emissions, and imposes heavier punitive policies on enterprises’ environmentally unfriendly actions, the enterprises will opt to choose the high-cost plan to improve their environmental performance.

References Agan Y, Acar MF, Borodin A (2013) Drivers of environmental processes and their impact on performance: a study of Turkish SMEs. J Clean Prod 51:23–33 Ahmad S, Tahar RM, Muhammady-Sukki F, Munir AB, Rahim RA (2015) Role of feed-in tariff policy in promoting solar photovoltaic investments in Malaysia: a system dynamics approach. Energy 84:808–815 Avinadav T, Herbon A, Spiegel U (2013) Optimal inventory policy for a perishable item with demand function sensitive to price and time. Int J Prod Econ 144:497–506 Babar M, Nyugen PH, Cuk V, Kamphuis IR, Bongaerts M, Hanzelka Z (2016) The rise of AGILE demand response: enabler and foundation for change. Renew Sust Energy Rev 56:686–693 Badcock J, Lenzen M (2010) Subsidies for electricity-generating technologies: a review. Energy Policy 38:5038–5047 Bi K, Huang P, Ye H (2015) Risk identification, evaluation and response of low-carbon technological innovation under the global value chain: a case of the Chinese manufacturing industry. Technol Forecast Soc Change 100:238–248 Bianco V, Manca O, Nardini S (2009) Electricity consumption forecasting in Italy using linear regression models. Energy 34:1413–1421 Blumberga A, Timma L, Romagnoli F, Blumberga D (2015) Dynamic modelling of a collection scheme of waste portable batteries for ecological and economic sustainability. J Clean Prod 88:224–233 Chen A, Groenewold N (2015) Emission reduction policy: a regional economic analysis for China. Econ Model 51:136–152 Chen YS (2008) The driver of green innovation and green image–green core competence. J Bus Ethics 81:531–543 China Economic Herald (2013) The incentive standard of energy performance contracting may be increased as 500 Yuan per tonne of standard coal. http://www.ceh.com.cn/ztbd/jnjpzk/277781. shtml. Accessed 14 Oct 2015

References

129

China HVACR (China Heating Ventilation and Air Conditioning Refrigeration) (2015) The popularization of air conditioner energy-saving technologies. http://www.chinahvacr.com/hyzxnews/ show-htm-itemid-3122648.html. Accessed 15 Jan 2016 Choi Y (2014) Global e-business management: theory and practice. Bomyoung-books Publishing Co, Seoul Choi Y (2015) Introduction to the special issue on “Sustainable e-governance in Northeast-Asia: challenges or sustainable innovation”. Technol Forecas Soc Change 96:1–3 Chua KJ, Chou SK, Yang WM, Yan J (2013) Achieving better energy-efficient air conditioning–a review of technologies and strategies. Appl Energy 104:87–104 Cornillie J, Fankhauser S (2004) The energy intensity of transition countries. Energy Econ 26:283– 295 Craig CA, Allen MW (2014) Enhanced understanding of energy ratepayers: factors influencing perceptions of government energy efficiency subsidies and utility alternative energy use. Energy Policy 66:224–233 Cressman R (2003) Evolutionary dynamics and extensive form games. MIT Press, Cambridge Dangelico RM, Pontrandolfo P (2010) From green product definitions and classifications to the green option matrix. J Clean Prod 18:1608–1628 Davis MD (1983) Game theory: a nontechnical introduction. Dover Publications, Mineola Dayaratne SP, Gunawardana KD (2015) Carbon footprint reduction: a critical study of rubber production in small and medium scale enterprises in Sri Lanka. J Clean Prod 103:87–103 Diamond D (2009) The impact of government incentives for hybrid-electric vehicles: evidence from US states. Energy Policy 37:972–983 Dijk M, Yarime M (2010) The emergence of hybrid-electric cars: innovation path creation through co-evolution of supply and demand. Technol Forecast Soc Chang 77:1371–1390 Dowlatshahi S (2005) Strategic success factors in enterprise resource-planning design and implementation: a case-study approach. Int J Prod Res 43:3745–3771 Dutta PK (1999) Strategies and games: theory and practice. MIT Press, Cambridge Elschner C, Ernst C, Licht G, Spengel C (2011) What the design of an R&D tax incentive tells about its effectiveness: a simulation of R&D tax incentives in the European Union. J Technol Transf 36:233–256 Feng LH, Zhang XC, Luo GY (2008) Application of system dynamics in analyzing the carrying capacity of water resources in Yiwu City, China. Math Comput Simulat 79:269–278 Friedman D (1991) Evolutionary games in economics. Econometrica 59:637–666 Galinato GI, Yoder JK (2010) An integrated tax-subsidy policy for carbon emission reduction. Resour Energy Econ 32:310–326 Geng Y, Doberstein B (2008) Greening government procurement in developing countries: building capacity in China. J Environ Manage 88:932–938 Gintis H (2009) Game theory evolving: a problem-centered introduction to modeling strategic interaction. Princeton University Press, Princeton Guenther M, Saunders CM, Tait PR (2012) Carbon labeling and consumer attitudes. Carbon Manag 3:445–455 Guo X, Guo X (2015) China’s photovoltaic power development under policy incentives: a system dynamics analysis. Energy 93:589–598 Harrison R, Jaumandreu J, Mairesse J, Peters B (2014) Does innovation stimulate employment? A firm-level analysis using comparable micro-data from four European countries. Int J Ind Organ 35:29–43 He Y, Zhang J (2015) Real-time electricity pricing mechanism in China based on system dynamics. Energy Convers Manage 94:394–405 Hsu CW (2012) Using a system dynamics model to assess the effects of capital subsidies and feed-in tariffs on solar PV installations. Appl Energy 100:205–217 Huang G (2006) The determinants of capital structure: evidence from China. China Econ Rev 17:14–36

130

3 Interaction Among Stakeholders Involved in Carbon Labeling Scheme

Hui EC, Bao H (2013) The logic behind conflicts in land acquisitions in contemporary China: a framework based upon game theory. Land Use Policy 30:373–380 Iles A, Martin AN (2013) Expanding bioplastics production: sustainable business innovation in the chemical industry. J Clean Prod 45:38–49 Janssen MA, Jager W (2002) Stimulating diffusion of green products. J Evol Econ 12:283–306 Ji P, Ma X, Li G (2015) Developing green purchasing relationships for the manufacturing industry: an evolutionary game theory perspective. Int J Prod Econ 166:155–162 Jin GR, Zhang L (2014) Efficiency of energy saving and emissions reduction in small and mediumsized enterprises and its influencing factors. China Soft Sci 1:126–133 (In Chinese) Jin W, Xu L, Yang Z (2009) Modeling a policy making framework for urban sustainability: Incorporating system dynamics into the Ecological Footprint. Ecol Econ 68:2938–2949 Johnstone ML, Tan LP (2015) Exploring the gap between consumers’ green rhetoric and purchasing behaviour. J Bus Ethics 132:311–328 Kanada M, Fujita T, Fujii M, Ohnishi S (2013) The long-term impacts of air pollution control policy: historical links between municipal actions and industrial energy efficiency in Kawasaki City, Japan. J Clean Prod 58:92–101 Kelly A (2003) Decision making using game theory: an introduction for manager. Cambridge University Press, Cambridge Kelly FP, Maulloo AK, Tan DK (1998) Rate control for communication networks: shadow prices, proportional fairness and stability. J Oper Res Soc 49:237–252 Kim DH, Kim DH (1997) System dynamics model for a mixed strategy game between police and driver. Syst Dyn Rev 13:33–52 Kocabiyikoglu A, Popescu I (2011) An elasticity approach to the newsvendor with price-sensitive demand. Oper Res 59:301–312 Kong LB, Price L, Hasanbeigi A, Liu HB, Li JG (2013) Potential for reducing paper mill energy use and carbon dioxide emissions through plant-wide energy audits: a case study in China. Appl Energy 102:1334–1342 Koufteros X, Vonderembse M, Jayaram J (2005) Internal and external integration for product development: the contingency effects of uncertainty, equivocality, and platform strategy. Decision Sci 36:97–133 Kreng VB, Wang BJ (2013) An innovation diffusion of successive generations by system dynamics—an empirical study of Nike Golf Company. Technol Forecast Soc Chang 80:77–87 Kyle P, Clarke L, Rong F, Smith SJ (2010) Climate policy and the long-term evolution of the US buildings sector. Energy J 31:145–172 Li J, Kendall G, John R (2016) Computing Nash equilibria and evolutionarily stable states of evolutionary games. IEEE Trans Evol Comput 20:460–469 Lin B, Jiang Z (2011) Estimates of energy subsidies in China and impact of energy subsidy reform. Energy Econ 33:273–283 Lin J, Rosenquist G (2008) Stay cool with less work: China’s new energy-efficiency standards for air conditioners. Energ Policy 36:1090–1095 Lin RJ, Tan KH, Geng Y (2013) Market demand, green product innovation, and firm performance: evidence from Vietnam motorcycle industry. J Clean Prod 40:101–107 Liu XB, Wang C, Zhang WS, Suk S, Sudo K (2013) Company’s affordability of increased energy costs due to climate policies: a survey by sector in China. Energ Econ 36:419–430 Lu C, Zhang X, He J (2010) A CGE analysis to study the impacts of energy investment on economic growth and carbon dioxide emission: a case of Shaanxi Province in western China. Energy 35:4319–4327 Luce RD, Raiffa H (1989) Games and decisions: introduction and critical survey. Dover Publications, New York Luo Y, Xue Q, Han B (2010) How emerging market governments promote outward FDI: experience from China. J World Bus 45:68–79

References

131

Mahlia TMI, Tohno S, Tezuka T (2013) International experience on incentive program in support of fuel economy standards and labelling for motor vehicle: a comprehensive review. Renew Sust Energ Rev 25:18–33 Michelsen O, de Boer L (2009) Green procurement in Norway: a survey of practices at the municipal and county level. J Environ Manag 91:160–167 Ministry of Industry and Information Technology of the People’s Republic of China (2012) Detailed rules to promote energy efficient room air conditioners. http://www.miit.gov.cn/n11293472/n11 293832/n11294042/n11302360/14771928.html. Accessed 15 Jul 2017 Myeong S, Kwon Y, Seo H (2014) Sustainable E-governance: the relationship among trust, digital divide, and E-government. Sustainability 6:6049–6069 Nash J (1951) Non-Cooperative Games. Ann Math 54:286–295 National Bureau of Statistics of the People’s Republic of China (2015) Sales of air-conditioners. http://data.stats.gov.cn/easyquery.htm?cn=B01. Accessed 14 Jul 2015 Ng WPQ, Lam HL, Ng FY, Kamal M, Lim JHE (2012) Waste-to-wealth: green potential from palm biomass in Malaysia. J Clean Prod 34:57–65 Ockwell DG, Watson J, MacKerron G, Pal P, Yamin F (2008) Key policy considerations for facilitating low carbon technology transfer to developing countries. Energy Policy 36:4104–4115 OECD (2016) Green growth and consumer behaviour. http://www.oecd.org/greengrowth/greeng rowthandconsumerbehaviour.htm. Accessed 16 Jul 2017 Olive H, Volschenk J, Smit E (2011) Residential consumers in the Cape Peninsula’s willingness to pay for premium priced green electricity. Energy Policy 39:544–550 Olson EL (2013) It’s not easy being green: the effects of attribute trade offs on green product preference and choice. J Acad Market Sci 41:171–184 Pan SY, Du MA, Huang IT, Liu IH, Chang EE, Chiang PC (2015) Strategies on implementation of waste-to-energy (WTE) supply chain for circular economy system: a review. J Clean Prod 108:409–421 Peterson DW, Eberlein RL (2006) Reality check: a bridge between systems thinking and system dynamics. Syst Dyn Rev 10:159–174 Qi C, Chang NB (2011) System dynamics modeling for municipal water demand estimation in an urban region under uncertain economic impacts. J Environ Manage 92:1628–1641 Rapoport A (1999) Two-person game theory. Mineola, New York Ramayah T, Lee JWC, Mohamad O (2010) Green product purchase intention: some insights from a developing country. Resour Conserv Recycl 54:1419–1427 Romp G (1997) Game theory introduction and application. Oxford University Press, Oxford Seroka-Stolka O (2016) Green initiatives in environmental management of logistics companies. Transp Res Procedia 16:483–489 Sgobbi A, Simões SG, Magagna D, Nijs W (2016) Assessing the impacts of technology improvements on the deployment of marine energy in Europe with an energy system perspective Renew. Energy 89:515–525 Shuai CM, Ding LP, Zhang YK, Guo Q, Shuai J (2014) How consumers are willing to pay for low-carbon products? Results from a carbon-labeling scenario experiment in China. J Clean Prod 83:366–373 Sloan TW (2011) Green renewal: incorporating environmental factors in equipment replacement decisions under technological change. J Clean Prod 19:173–186 Soldaat L, Visser H, van Roomen M, van Strien A (2007) Smoothing and trend detection in waterbird monitoring data using structural time-series analysis and the Kalman filter. J Ornithol 148:351– 357 Standardization Administration of the People’s Republic of China (2010) The minimum allowable value of the energy efficiency and energy efficiency grades for room air conditioners. http://www. docin.com/p-90921485.html. Accessed 12 Oct 2015 (In Chinese) Sterman JD (2000) Business dynamics: systems thinking and modeling for a complex world. New York, US Straffin PD (1993) Game theory and strategy. Washington, US

132

3 Interaction Among Stakeholders Involved in Carbon Labeling Scheme

Su J, He J (2010) Does giving lead to getting? Evidence from Chinese private enterprises. J Bus Ethics 93:73–90 Suk S, Liu XB, Lee SY, Go S, Sudo K (2014) Affordability of energy cost increases for Korean companies due to market-based climate policies: a survey study by sector. J Clean Prod 67:208– 219 Suk S, Liu XB, Sudo K (2013) A survey study of energy saving activities of industrial companies in the Republic of Korea. J Clean Prod 41:301–311 Suryani E, Chou SY, Hartono R, Chen CH (2010) Demand scenario analysis and planned capacity expansion: a system dynamics framework. Simul Model Pract Theory 18:732–751 Tan LP, Johnstone ML, Yang L (2016) Barriers to green consumption behaviors: the role of consumers’ green perceptions. Austra J Market 24:288–299 Tan M, Tan R, Khoo H (2014) Prospects of carbon labelling-a life cycle point of view. J Clean Prod 72:76–88 Tang JR, Hao XD, Zhang BY (2012) Control of carbon emissions intensity based on a system dynamics approach. Stat Dec 9:63–65 (In Chinese) Tang L, Wu JQ, Yu L, Bao Q (2015) Carbon emissions trading scheme exploration in China: a multi-agent-based model. Energy Policy 81:152–169 The Central People’s Government of the People’s Republic of China (2007) Enterprise income tax law and implementation rules of the People’s Republic of China. http://www.gov.cn/flfg/200703/19/content_554243.htm. Accessed 20 Jul 2015 Thomas LC (2003) Games, theory, and applications. Mineola, US Thongplew N, van Koppen CSA, Spaargaren G (2014) Companies contributing to the greening of consumption: findings from the dairy and appliance industries and Thailand. J Clean Prod 75:96–105 Tian Y, Govindan K, Zhu Q (2014) A system dynamics model based on evolutionary game theory for green supply chain management diffusion among Chinese manufacturers. J Clean Prod 80:96–105 Tsou YS, Wang HF (2012) Subsidy and penalty strategy for a green industry sector by bi-level mixed integer nonlinear programming. J Chin Inst Ind Eng 29:226–236 Upham P, Dendler L, Bleda M (2011) Carbon labelling of grocery products: public perceptions and potential emissions reductions. J Clean Prod 19:348–355 US Environmental Protection Agency (2010) Available and emerging technologies for reducing greenhouse gas emissions from the pulp and paper manufacturing industry. https://www.epa.gov/ sites/production/files/2015-12/documents/pulpandpaper.pdf. Accessed 18 Oct 2014 Wang P, Song Y (2010) Technological guideline of carbon emissions reduction and energy saving. Beijing, China (In Chinese) Wang Y, Chang X, Chen Z, Zhong Y, Fan T (2014) Impact of subsidy policies on recycling and remanufacturing using system dynamics methodology: a case of auto parts in China. J Clean Prod 74:161–171 Wu JH, Huang YL, Liu CC (2011) Effect of floating pricing policy: an application of system dynamics on oil market after liberalization. Energy Policy 39:4235–4252 Yunna W, Kaifeng C, Yisheng Y, Tiantian F (2015) A system dynamics analysis of technology, cost and policy that affect the market competition of shale gas in China. Renew Sust Energy Rev 45:235–243 Yusup MZ, Mahmood WHW, Salleh MR, Ab Rahman MN (2015) The implementation of cleaner production practices from Malaysian manufacturers’ perspectives. J Clean Prod 108:659–672 Zeng SX, Meng XH, Yin HT, Tam CM, Sun L (2010) Impact of cleaner production on business performance. J Clean Prod 18:975–983 Zhang C, He W, Hao R (2016) Comprehensive estimation of the financial risk of iron and steel enterprise-based on carbon emission reduction. J Sci Ind Res India 75:143–149 Zhang X, Srinivasan R, Bosch D (2009) Calibration and uncertainty analysis of the SWAT model using genetic algorithms and Bayesian model averaging. J Hydrol 374:07–317

References

133

Zhang YJ, Wang AD, Tan W (2015) The impact of China’s carbon allowance allocation rules on the product prices and emission reduction behaviors of ETS-covered enterprises. Energy Policy 86:176–185 Zhao R, Peng D, Li, Y (2015) An interaction between government and manufacturer in implementation of cleaner production: a multi-stage game theoretical analysis. Int J Environ Res 9(3):1069-1078 Zhao R, Deutz P, Neighbour G, McGuire M (2012a) Carbon emissions intensity ratio: an indicator for an improved carbon labelling scheme. Environ Res Lett 7:9 Zhao R, Neighbour G, Han J, McGuire M, Deutz P (2012b) Using game theory to describe strategy selection for environmental risk and carbon emissions reduction in the green supply chain. J Loss Prevent Proc 25:927–936 Zhao R, Neighbour G, McGuire M, Deutz P (2013) A software based simulation for cleaner production: a game between manufacturers and government. J Loss Prevent Proc. 26:59–67 Zhao R, Zhong S (2015) Carbon labelling influences on consumers’ behaviour: a system dynamics approach. Ecol Indic 51:98–106 Zhao R, Zhou X, Han J, Liu C (2016) For the sustainable performance of the carbon reduction labeling policies under an evolutionary game simulation. Technol Forecast Soc Change 112:262– 274 Zhao R, Zhou X, Jin Q, Wang Y, Liu CL (2017) Enterprises’ compliance with government carbon reduction labelling policy using a system dynamics approach. J Clean Prod 163:303–319 Zhou KZ, Brown JR, Dev CS (2009) Market orientation, competitive advantage, and performance: a demand-based perspective. J Bus Res 62:1063–1070 Zhu Q, Geng Y, Sarkis J (2013) Motivating green public procurement in China: an individual level perspective. J Environ Manage 126:85–95 Zhu Q, Sarkis J (2006) An inter-sectoral comparison of green supply chain management in China: drivers and practices. J Clean Prod 14:472–486 ZOL (2015) The annual reports of air conditioner market in China for 2014–2015. http://tech.hexun. com/2015-02-05/173107490.html. Accessed 15 Jul 2015.

Chapter 4

Carbon Labeling Improvement and Its Application

Abstract The carbon emissions embodied in a product are usually presented to consumers in the form of a carbon label on product packaging. However, the current carbon labelling system does not communicate a sufficiently meaningful message to the consumer. How to improve transparency in carbon labeling schemes to provide sufficient information is significant to drive sustainable production and consumption. This Chapter proposes an improved carbon labelling scheme, and explores whether such a carbon labeling system may be applied to benchmarking a low carbon community, to develop policy implications for the creation of low-carbon lifestyles. Keywords Carbon labeling scheme · Communication · Transparency · Benchmarking

4.1 Introduction The carbon emissions embodied in a product are usually presented to consumers in the form of a carbon label on product packaging. Over the past few years, managers have been encouraged to measure and report lifecycle-based carbon emissions (e.g., by using carbon reduction labeling schemes to incentivize emissions reductions and energy saving; Carbon Trust 2008; Weidema et al. 2008). However, the complexities of lifecycle assessment and difficulties obtaining a globally recognized protocol for standardized carbon footprint methodologies limit the accuracy of the calculated values (Cohen and Vandenbergh 2012; Hetherington et al. 2014). For example, the main standards related to carbon footprinting are PAS 2050, ISO 14,067, and WRI/WBCSD, which have different system boundaries and may give rise to varied carbon footprint information (Wu et al. 2014). This further complicates consumers’ decision making, as the product shows the same level of quality but is labeled with different CO2 values. For instance, based upon the responses from focus groups and surveys of responses to carbon labelling, it is clear that the public find it quite difficult to imagine a given quantity of CO2 emission and its potential environmental impact (Upham et al. 2011; Gadema and Oglethorpe 2011). It is implied that the current system of carbon labelling does not communicate a sufficiently meaningful message to the consumer. How to improve transparency in carbon reduction labeling schemes © Springer Nature Singapore Pte Ltd. 2021 R. Zhao and Y. Geng, Carbon Labeling Practice, https://doi.org/10.1007/978-981-16-2583-1_4

135

136

4 Carbon Labeling Improvement and Its Application

to provide sufficient information to consumers affects the low carbon consumption significantly (Harbaugh et al. 2011; Wu et al. 2014). Certified carbon labeling can not only provide information communication between producers and consumers but also bring more sustainable and ethical business practices (Zhu et al. 2013). Stewardship or supervision by both the national and local government is critical to promote low carbon consumption. Without a credible evaluation system, consumers may not buy low carbon products even with such intentions (Tanner 2006). In addition, this should be done by considering the local situations. Due to the different cultural contexts, development stage and challenges that difference regions are facing, it would be rational for different activities by considering their own realities. In such context, the Chapter is divided into three Sections. Section 4.1 proposes a new carbon labelling scheme by normalizing carbon emissions data on a common indicator ‘carbon emissions intensity ratio’ to improve the visibility of products’ life cycle carbon emissions. Section 4.2 further proposes a carbon emission reduction label for benchmarking and classifying industries/sectors in terms of their emission reduction responsibilities. Section 4.3 explores whether such a carbon labeling system may be applied to benchmarking a low carbon community, to develop policy implications for the creation of low-carbon lifestyles.

4.2 An Improved Carbon Labelling Scheme This section proposes a dimensionless system of carbon emissions labelling, which, by facilitating comparison between products, would increase the ability of green consumers to implement their concerns in the market place. The carbon emissions data is normalized to a common scale of Carbon Emissions Intensity (CEI), and a new indicator Carbon Emissions Intensity Ratio is generated based the ratio of CEI to the annual national greenhouse gas emission per gross domestic product. The value of the dimensionless Carbon Emissions Intensity Ratio (CEIR) of a product can be evaluated on a simple scale with five ranges of values from ‘extremely low’ to ‘extremely high’. The performance of a given product can be presented on its packaging by a simple diagram with colour gradation. It is hoped that this study could lead to an improvement in the clarity of current carbon labelling schemes and thus encourage consumers to select low carbon products, as well as increasing the potential for reduction in carbon emissions. The approach used in this study emphasises the ease with which the consumer can apply the information. Thus, the indicator developed is dimensionless to aid comprehension by the public and comparability of values between products. Furthermore, it is presented in a simple, visually appealing manner. Nonetheless, the calculations behind the proposed indicator, whilst mathematically straightforward, are more complex than a typical consumer is likely to want to comprehend. This

4.2 An Improved Carbon Labelling Scheme

137

underlying complexity is true of LCA itself, but the proposed methodology has the advantage of producing a dimensionless number, which is widely comparable and which can be presented as a simple graphic. This study outlines and demonstrates the mechanics of the process for policy makers/companies, as well as presenting sample labels.

4.2.1 Method A carbon footprint derived from the LCA is generally determined by functional unit. That is the carbon emissions relate to a specific scenario, e.g. per pack, per serving, per pint etc. (Carbon Trust 2010). The starting point of the methodology is to normalize the carbon footprint into a common scale. We suggest using an indicator defined as ‘Carbon Emissions Intensity’ (CEI), which can be understood as carbon emissions per unit of economic output (DEFRA 2009). It can be calculated as follows: Carbon Emissions Intensity (CEIi ) =

CEi Ri ( j)

(4.1)

where CEi is the carbon emissions of the ith product (kilogram per functional unit), which is derived from a LCA; Ri (j) is the retail price of the ith product at the the jth year, using British pound per functional unit. However, the CEI is highly dependent upon the retailing price, fluctuations of which would disguise temporal variations in carbon emissions levels (DEFRA 2009). Thus, in successive years the product’s retailing price would need to be adjusted to allow for inflation. This is not a trivial matter, but the UK government, for example, provides guidance on the derivation of official measures of inflation, as well as tracking the calculated values (Office of National Statistics 2012). The second stage is to devise a baseline to build a dimensionless indicator and frame of reference. The baseline is the national carbon emissions per unit of gross domestic product (GDP), defined as National Carbon Emissions Intensity (NCEI) (Fan et al. 2007; Wang et al. 2011). This is expressed as follows: NCEI( j) =

GHG( j) GDP( j)

(4.2)

where NCEI(j) is the National Carbon Emissions Intensity for the country of production in a designated year (j); GHG(j) is the national greenhouse gas emissions (direct emissions) at the jth year and GDP(j) is the national gross domestic product at the jth year. It is noted that NCEI is based on estimated emissions for a given year, rather than life cycle emissions for a product. This is appropriate since establishing an emissions intensity baseline in time, to facilitate comparison of product emissions intensity over time, as well as between products at a given time.

138

4 Carbon Labeling Improvement and Its Application

From Eqs. (4.1) and (4.2), a dimensionless indicator is set up based upon the ratio of CEI and NCEI, which can be defined as ‘Carbon Emissions Intensity Ratio’ (CEIR) and expressed as follows: CEIRi =

CEIi = NCEI( j)

CEi Ri ( j) GHG( j) GDP( j)

=

GDP( j) CEi × Ri ( j) GHG( j)

(4.3)

By definition, the value for NCEI for the designated baseline year would remain constant for the calculation of CEIR for successive years. Thus any change in CEIR over time would be accounted for primarily by changes in carbon emissions per product unit for the given product. It is assumed that CEIRs are normally distributed with mean value μ and standard deviation ± σ. This assumption may be compromised if, for example, products for which data are available are preferentially from higher emissions categories. However, this potential problem would decrease over time, as labelling became more widely adopted. Based on this assumption, it is suggested dividing the carbon emissions intensity ratio into five ranges, designated as extremely low, low, medium, high and extremely high, respectively (see Fig. 4.1). Both μ and σ are calculated based on Eq. (4.3), in order to determine the regional boundaries. Thus, the mean of carbon emissions intensity ratio μ can be expressed as follows: n CEIRi (4.4) μ = i=1 n

Fig. 4.1 Ranges of carbon emissions intensity ratio in terms of the normal distribution

4.2 An Improved Carbon Labelling Scheme

Extremely high

139

r>μ+σ

High

μ+σ/2