上海交通大学：公共管理类《专业英语》课程教学资源（阅读材料）A Game Theory Based Analysis of Decision Making for Green Retrofit under Different Occupancy Types

Journal of Cleaner Production 137(2016)1300-1312 Contents lists available at ScienceDirect Cleaner Production Journal of Cleaner Production ELSEVIER journal homepage:www.elsevier.com/locate/jclepro A game theory based analysis of decision making for green retrofit CrossMark under different occupancy types Xin Liang Yi Peng,Geoffrey Qiping Shen . School of International and Public Affairs,Shanghai Jiao Tong University.Shanghai.China bDepartment of Building and Real Estate.Hong Kong Polytechnic University.Hong Kong China School of Public Administration,Zhejiang University of Finance Economics,Hangzhou,China ARTICLE INFO ABSTRACT Article history: Buildings are responsible for almost half of all the energy consumption and greenhouse gas(GHG) Received 13 October 2014 emissions in the world.This situation highlights the importance of the green retrofit for existing Received in revised form buildings in reducing the energy consumption and GHG emissions,as emphasized by the academia and 5 April 2016 Accepted 28 July 2016 improved by the government.Although relevant stakeholders are interested in implementing green Available online 3 August 2016 retrofit projects,this approach has not been widely pursued by the industry.the reasons of which remain unclear.Therefore,this study aims to reveal the underlying logic by analyzing the behaviors of the Keywords: building owners and occupiers,who are the direct decision makers in initiating green retrofit at the Green retrofit initial intention phase.Three occupancy scenarios,namely,owner-occupied(baseline scenario),single- Decision making occupied,and multi-occupied buildings,are used for the game analysis.The Nash Equilibrium of the Game theory game is used to analyze the probable decisions of the owners and occupiers under the last two occu- Owner pancy scenarios.Results demonstrate that both owners and occupiers are reluctant to retrofit under both Occupier scenarios.Nevertheless,the reasons vary under the two scenarios despite the same results obtained.This study clarifies the reasons for the reluctance of the direct decision makers to participate in green retrofit projects.The main reasons include the split incentives between the owners and occupiers,the complex coordination,and the uncertainty of green retrofit.The identified reasons are also beneficial to the policy makers,particularly in their effort to promote green retrofit by considering the requirements of owners and occupiers under the different occupancy types. 2016 Elsevier Ltd.All rights reserved. 1.Introduction and economy (Menassa and Baer.2014).In the life cycle of a building,more than 80%of the energy consumption occurs during Buildings are responsible for the majority of the energy con- the actual occupancy operation stage rather than during the con- sumption and greenhouse gas(GHG)emissions around the world. struction stage (UNEP,2007). In the United States(US).buildings consume approximately 50%of Owing to its essential influence on energy consumption and the total energy(ElA,2010),while in Europe,the ratio is approxi- GHG emission,the green retrofit for existing buildings should be mately 40%(Kashif et al.,2011).In the last few decades,the building given due attention in relation to sustainable development."Green energy consumption has continuously increased,particularly in retrofit"can be defined as the incremental improvement of the developing countries.In China,the building energy consumption fabric and systems of a building with the primary intention of increased by more than 10%annually(Xu et al.,2011).Buildings are improving energy efficiency and reducing carbon emissions.It can also considered responsible for a quarter of the total global carbon likewise refer to other terms in literature,such as refurbishment. dioxide (CO2)emissions (Hong et al.,2015).which consequently rehabilitation,modernization,renovation,improvements,adapta- increase the adverse impact on the global environment,healthcare, tion,additions,repairs and renewal on existing buildings(Ali and Rahmat,2009).However,routine maintenance and cleaning work are excluded (Quah,1988). Green retrofit for existing buildings is emphasized by govern- Corresponding author. E-mail addresses:liangxinpku@gmail.com (X.Liang).pengyihz@gmail.com ments all over the world.The US government passed the Energy (Y.Peng).geoffrey.shen@polyu.edu.hk(G.QShen) Policy Act (EPA)of 2005 and the Executive Order 13423,which http://dx.doi.org/10.1016/j.jclepro.2016.07.200 0959-6526/0 2016 Elsevier Ltd.All rights reserved

A game theory based analysis of decision making for green retrofit under different occupancy types Xin Liang a, b , Yi Peng c , Geoffrey Qiping Shen b, * a School of International and Public Affairs, Shanghai Jiao Tong University, Shanghai, China b Department of Building and Real Estate, Hong Kong Polytechnic University, Hong Kong, China c School of Public Administration, Zhejiang University of Finance & Economics, Hangzhou, China article info Article history: Received 13 October 2014 Received in revised form 5 April 2016 Accepted 28 July 2016 Available online 3 August 2016 Keywords: Green retrofit Decision making Game theory Owner Occupier abstract Buildings are responsible for almost half of all the energy consumption and greenhouse gas (GHG) emissions in the world. This situation highlights the importance of the green retrofit for existing buildings in reducing the energy consumption and GHG emissions, as emphasized by the academia and improved by the government. Although relevant stakeholders are interested in implementing green retrofit projects, this approach has not been widely pursued by the industry, the reasons of which remain unclear. Therefore, this study aims to reveal the underlying logic by analyzing the behaviors of the building owners and occupiers, who are the direct decision makers in initiating green retrofit at the initial intention phase. Three occupancy scenarios, namely, owner-occupied (baseline scenario), single￾occupied, and multi-occupied buildings, are used for the game analysis. The Nash Equilibrium of the game is used to analyze the probable decisions of the owners and occupiers under the last two occu￾pancy scenarios. Results demonstrate that both owners and occupiers are reluctant to retrofit under both scenarios. Nevertheless, the reasons vary under the two scenarios despite the same results obtained. This study clarifies the reasons for the reluctance of the direct decision makers to participate in green retrofit projects. The main reasons include the split incentives between the owners and occupiers, the complex coordination, and the uncertainty of green retrofit. The identified reasons are also beneficial to the policy makers, particularly in their effort to promote green retrofit by considering the requirements of owners and occupiers under the different occupancy types. © 2016 Elsevier Ltd. All rights reserved. 1. Introduction Buildings are responsible for the majority of the energy con￾sumption and greenhouse gas (GHG) emissions around the world. In the United States (US), buildings consume approximately 50% of the total energy (EIA, 2010), while in Europe, the ratio is approxi￾mately 40% (Kashif et al., 2011). In the last few decades, the building energy consumption has continuously increased, particularly in developing countries. In China, the building energy consumption increased by more than 10% annually (Xu et al., 2011). Buildings are also considered responsible for a quarter of the total global carbon dioxide (CO2) emissions (Hong et al., 2015), which consequently increase the adverse impact on the global environment, healthcare, and economy (Menassa and Baer, 2014). In the life cycle of a building, more than 80% of the energy consumption occurs during the actual occupancy operation stage rather than during the con￾struction stage (UNEP, 2007). Owing to its essential influence on energy consumption and GHG emission, the green retrofit for existing buildings should be given due attention in relation to sustainable development. “Green retrofit” can be defined as the incremental improvement of the fabric and systems of a building with the primary intention of improving energy efficiency and reducing carbon emissions. It can likewise refer to other terms in literature, such as refurbishment, rehabilitation, modernization, renovation, improvements, adapta￾tion, additions, repairs and renewal on existing buildings (Ali and Rahmat, 2009). However, routine maintenance and cleaning work are excluded (Quah, 1988). Green retrofit for existing buildings is emphasized by govern￾ments all over the world. The US government passed the Energy Policy Act (EPA) of 2005 and the Executive Order 13423, which * Corresponding author. E-mail addresses: liangxinpku@gmail.com (X. Liang), pengyihz@gmail.com (Y. Peng), geoffrey.shen@polyu.edu.hk (G.Q. Shen). Contents lists available at ScienceDirect Journal of Cleaner Production journal homepage: www.elsevier.com/locate/jclepro http://dx.doi.org/10.1016/j.jclepro.2016.07.200 0959-6526/© 2016 Elsevier Ltd. All rights reserved. Journal of Cleaner Production 137 (2016) 1300e1312

X.Liang et aL Journal of Cleaner Production 137 (2016)1300-1312 1301 require retrofitting 15%of the total number of the existing buildings a potential and ongoing green retrofit project (Gucyeter and to improve energy efficiency by 2015 compared with the 2003 Gunaydin,2012).The main stakeholders identified by the litera- baseline (EPA,2005).Approximately 30 billion US dollars are ear- ture include the client/owner,occupier/tenant,facilities manager, marked to conduct green retrofit projects for the existing buildings consultant/designer,contractor,subcontractor,supplier,govern- and facilities.In 2010,the UK government launched the Carbon ment,financial institutions,energy service companies,environ- Reduction Commitment(CRC)Energy Efficiency Scheme to save 1.2 mental organization,professional association,media,public,labor million tons of CO,emissions annually by 2020.This scheme mo- union,and researcher/educator (Gultekin et al.,2013:Juan et al. tivates the consumers to consider energy efficiency options and to 2009:Kaklauskas et al.,2004.2008:Miller and Buys,2008;Yang invest in building retrofit projects.Similarly,the Chinese govern- and Zou,2014).which are shown in Fig.1.Previous studies asser- ment has introduced various policies.Chinese 12th Five-Year Plan ted that the process of green retrofit projects can normally be stipulated that 400 million m2 residential buildings and 60 million divided into five phases,namely.the 1)initial intention or setup,2) m public buildings are planned to be retrofitted as pilot projects pre-retrofit survey and energy performance assessment,3)design between 2011 and 2015 to improve building energy efficiency. 4)site implementation,and 5)validation and verification(Lapinski By contrast,green retrofit projects remain inadequately pursued et al.,2006;Ma et al.,2012).Various stakeholders are involved in in the industries.After the 2008 global economic recession,this green retrofit projects in the different phase (see Fig.1).For situation was further exacerbated by the challenge of ensuring the example,the energy consultants are normally involved in the pre- financial support for retrofit activities (Menassa,2011).Some pilot retrofit survey and energy performance assessment phase,whereas studies have revealed that the industries are unenthusiastic about the designers and contractors participate in the project at the green retrofit primarily because of the following aspects:the highly design and implementation phases. complex design analysis and solution (Davies and Osmani,2011: The owners and occupiers play important roles in making green Kasivisvanathan et al.,2012:Lapinski et al.,2006).intense inter- retrofit decisions,particularly at the very early stage,namely,initial disciplinary collaboration (Korkmaz et al.,2010:Lapinski et al. intention or setup phase (Liang et al..2015).In this phase,normally 2006),long payback periods (Kasivisvanathan et al.,2012: only the owners and occupiers are involved,who propose pre- Menassa,2011).financial problem (e.g.,limited access to capital, liminary retrofit plans and exchange opinions regarding retrofit. high cost,etc.)(Davies and Osmani,2011:Kasivisvanathan et al.. These stakeholders can decide whether to launch a retrofit project 2012:Xu et al,2011),lack of retrofit experience (Ali et al,2008: and to continue to the next steps of energy audit,design,and Kasivisvanathan et al.,2012:Korkmaz et al.,2010),and lack of implementation.The important role of owners in green retrofit can understanding of the available retrofit technologies (Davies and be naturally and easily understood,whereas the role of the occu- Osmani,2011:Miller and Buys,2008).Most of these research piers is often underestimated (Karvonen,2013).Juan et al.(2009) findings were obtained by analyzing the problem from technical, indicated that the influence of the occupiers makes the retrofit economic,and environmental perspectives.Only a few studies have more difficult and risky than new buildings because cooperation explored the behaviors of the main stakeholders,who may directly and participation of occupiers are required in an existing building decide whether a building retrofit can be implemented.In practice, retrofit (Miller and Buys,2008).In new buildings,the clients,who the owners and occupiers are the critical direct stakeholders in will become the building owners after construction,can decide by green retrofit at the initial intention phase.However,these in- themselves,whereas in retrofit,the owners have to consider oc- dividuals may have varying and conflicting opinions on whether a cupiers because of their lease contracts.The satisfaction of the building should be retrofitted and when and how the retrofit will occupiers can directly influence the occupancy rate,rent,and be implemented.Few studies,if not none,have investigated the owner reputation in the future.In addition to economic influence. decision behaviors of the occupiers and owners under different the actions of the occupiers are identified as major determinants of interaction relationships.Therefore,the logic for reluctance to energy consumption(Azar and Menassa,2012.2014).The occupiers conduct green retrofit activities in the industry remains unknown. can affect the energy consumption difference by up to 100% This study aims to reveal the underlying logic of the industry's through different behaviors,such as ventilation habits,indoor reluctance to conduct green retrofit by analyzing the behaviors of temperature setting behavior,and after-hour lighting use (Urge- owners and occupiers at the initial phase.This study differs from Vorsatz et al.,2009).Consequently.the occupiers are another the previous ones because instead of identifying the willingness of essential stakeholders in green retrofit projects.Numerous owners the stakeholders or the retrofit-related problems through a survey. and occupiers intend to carry out green retrofit,but,most of them it explores the behaviors of the direct decision makers under the are interrupted at the beginning because a consensus cannot be current market constraints through a game analysis.Modeling the reached.Ma et al.(2012)presented that the phase of deciding retrofit decision behaviors of the stakeholders under different whether to retrofit buildings,which is before the design phase,is scenarios with game theory can be an efficient means to properly the key phase of a sustainable building retrofit.Thus,answering the identify the underlying logic.The rest of the paper is organized into question of whether or not to retrofit a building by analyzing the eight sections.Section 2 critically reviews the literature on stake- interrelations between the owners and occupiers at the first phase holders in green retrofit and the relevant motivations and barriers. is fundamental. Section 3 describes the research methodology.Sections 4,5.and 6 The conventional studies related to the stakeholder analysis in introduce the specific game analysis in the different scenarios of green retrofit mainly focused on the owners and designers involved occupancy types.Section 7 comprehensively discusses the analytic in the energy assessment and design phase (Ali et al.,2008:Stiess results.Section 8 concludes the study and presents recommenda- and Dunkelberg.2013).However,a few recent studies have tions for future research examined the occupiers of existing buildings and their relationship with owners.Stephan and Menassa(2013)proposed an agent- 2.Literature review based model to analyze the social network interactions among the stakeholders(i.e.,owner,occupier,architect,and contractor)of 2.1.Critical stakeholders in green retrofit commercial buildings.In their subsequent study.Stephan and Menassa (2014)emphasized that the network structure and the The stakeholders in green retrofit are the people who,directly or confidence level of the stakeholders could significantly influence indirectly,have vested interests in a building and in the outcome of their own alignment toward a unified retrofit objective.This agent-

require retrofitting 15% of the total number of the existing buildings to improve energy efficiency by 2015 compared with the 2003 baseline (EPA, 2005). Approximately 30 billion US dollars are ear￾marked to conduct green retrofit projects for the existing buildings and facilities. In 2010, the UK government launched the Carbon Reduction Commitment (CRC) Energy Efficiency Scheme to save 1.2 million tons of CO2 emissions annually by 2020. This scheme mo￾tivates the consumers to consider energy efficiency options and to invest in building retrofit projects. Similarly, the Chinese govern￾ment has introduced various policies. Chinese 12th Five-Year Plan stipulated that 400 million m2 residential buildings and 60 million m2 public buildings are planned to be retrofitted as pilot projects between 2011 and 2015 to improve building energy efficiency. By contrast, green retrofit projects remain inadequately pursued in the industries. After the 2008 global economic recession, this situation was further exacerbated by the challenge of ensuring the financial support for retrofit activities (Menassa, 2011). Some pilot studies have revealed that the industries are unenthusiastic about green retrofit primarily because of the following aspects: the highly complex design analysis and solution (Davies and Osmani, 2011; Kasivisvanathan et al., 2012; Lapinski et al., 2006), intense inter￾disciplinary collaboration (Korkmaz et al., 2010; Lapinski et al., 2006), long payback periods (Kasivisvanathan et al., 2012; Menassa, 2011), financial problem (e.g., limited access to capital, high cost, etc.) (Davies and Osmani, 2011; Kasivisvanathan et al., 2012; Xu et al., 2011), lack of retrofit experience (Ali et al., 2008; Kasivisvanathan et al., 2012; Korkmaz et al., 2010), and lack of understanding of the available retrofit technologies (Davies and Osmani, 2011; Miller and Buys, 2008). Most of these research findings were obtained by analyzing the problem from technical, economic, and environmental perspectives. Only a few studies have explored the behaviors of the main stakeholders, who may directly decide whether a building retrofit can be implemented. In practice, the owners and occupiers are the critical direct stakeholders in green retrofit at the initial intention phase. However, these in￾dividuals may have varying and conflicting opinions on whether a building should be retrofitted and when and how the retrofit will be implemented. Few studies, if not none, have investigated the decision behaviors of the occupiers and owners under different interaction relationships. Therefore, the logic for reluctance to conduct green retrofit activities in the industry remains unknown. This study aims to reveal the underlying logic of the industry's reluctance to conduct green retrofit by analyzing the behaviors of owners and occupiers at the initial phase. This study differs from the previous ones because instead of identifying the willingness of the stakeholders or the retrofit-related problems through a survey, it explores the behaviors of the direct decision makers under the current market constraints through a game analysis. Modeling the retrofit decision behaviors of the stakeholders under different scenarios with game theory can be an efficient means to properly identify the underlying logic. The rest of the paper is organized into eight sections. Section 2 critically reviews the literature on stake￾holders in green retrofit and the relevant motivations and barriers. Section 3 describes the research methodology. Sections 4, 5, and 6 introduce the specific game analysis in the different scenarios of occupancy types. Section 7 comprehensively discusses the analytic results. Section 8 concludes the study and presents recommenda￾tions for future research. 2. Literature review 2.1. Critical stakeholders in green retrofit The stakeholders in green retrofit are the people who, directly or indirectly, have vested interests in a building and in the outcome of a potential and ongoing green retrofit project (Gucyeter and Gunaydin, 2012). The main stakeholders identified by the litera￾ture include the client/owner, occupier/tenant, facilities manager, consultant/designer, contractor, subcontractor, supplier, govern￾ment, financial institutions, energy service companies, environ￾mental organization, professional association, media, public, labor union, and researcher/educator (Gultekin et al., 2013; Juan et al., 2009; Kaklauskas et al., 2004, 2008; Miller and Buys, 2008; Yang and Zou, 2014), which are shown in Fig. 1. Previous studies asser￾ted that the process of green retrofit projects can normally be divided into five phases, namely, the 1) initial intention or setup, 2) pre-retrofit survey and energy performance assessment, 3) design, 4) site implementation, and 5) validation and verification (Lapinski et al., 2006; Ma et al., 2012). Various stakeholders are involved in green retrofit projects in the different phase (see Fig. 1). For example, the energy consultants are normally involved in the pre￾retrofit survey and energy performance assessment phase, whereas the designers and contractors participate in the project at the design and implementation phases. The owners and occupiers play important roles in making green retrofit decisions, particularly at the very early stage, namely, initial intention or setup phase (Liang et al., 2015). In this phase, normally only the owners and occupiers are involved, who propose pre￾liminary retrofit plans and exchange opinions regarding retrofit. These stakeholders can decide whether to launch a retrofit project and to continue to the next steps of energy audit, design, and implementation. The important role of owners in green retrofit can be naturally and easily understood, whereas the role of the occu￾piers is often underestimated (Karvonen, 2013). Juan et al. (2009) indicated that the influence of the occupiers makes the retrofit more difficult and risky than new buildings because cooperation and participation of occupiers are required in an existing building retrofit (Miller and Buys, 2008). In new buildings, the clients, who will become the building owners after construction, can decide by themselves, whereas in retrofit, the owners have to consider oc￾cupiers because of their lease contracts. The satisfaction of the occupiers can directly influence the occupancy rate, rent, and owner reputation in the future. In addition to economic influence, the actions of the occupiers are identified as major determinants of energy consumption (Azar and Menassa, 2012, 2014). The occupiers can affect the energy consumption difference by up to 100% through different behaviors, such as ventilation habits, indoor temperature setting behavior, and after-hour lighting use (Ürge￾Vorsatz et al., 2009). Consequently, the occupiers are another essential stakeholders in green retrofit projects. Numerous owners and occupiers intend to carry out green retrofit, but, most of them are interrupted at the beginning because a consensus cannot be reached. Ma et al. (2012) presented that the phase of deciding whether to retrofit buildings, which is before the design phase, is the key phase of a sustainable building retrofit. Thus, answering the question of whether or not to retrofit a building by analyzing the interrelations between the owners and occupiers at the first phase is fundamental. The conventional studies related to the stakeholder analysis in green retrofit mainly focused on the owners and designers involved in the energy assessment and design phase (Ali et al., 2008; Stiess and Dunkelberg, 2013). However, a few recent studies have examined the occupiers of existing buildings and their relationship with owners. Stephan and Menassa (2013) proposed an agent￾based model to analyze the social network interactions among the stakeholders (i.e., owner, occupier, architect, and contractor) of commercial buildings. In their subsequent study, Stephan and Menassa (2014) emphasized that the network structure and the confidence level of the stakeholders could significantly influence their own alignment toward a unified retrofit objective. This agent￾X. Liang et al. / Journal of Cleaner Production 137 (2016) 1300e1312 1301

1302 X.Liang et aL Journal of Cleaner Production 137(2016)1300-1312 Phases and Processes Internal External Stakeholders Stakeholders Propose intentions of retrofit; Define objectives and expectations; Owner. Initial intention Exchange attitudes towards retrofit. Occupier. setup Government <Retrofit?No End Policy: 2 IYes Standard: Pre-retrofit survey: Energy Financial Pre-retrofit survey Data collection and energy consultant; institution/bank and energy commitment; performance assessment: Facility performance Goal and target establishment. manager Financial assessment Support; Energy service <Confirmed?N End company Yes Energy service: Quantify retrofit measures: Cost-benefit analysis; Influence NGO: Academic Develop action plans; Designer Environmental Design Client review and comments. report; organization Technology <Satisfied?No development: Researcher Yes Contractor; Media; Educator Site Implementation and commissioning Sub-contractor; implementation Supplier. Public 5 Post-retrofit survey; pressure; Professional Retrofit report; Sustainable association Occupier satisfaction evaluation; Validation and development Client review and comments. verification promotion Surveyor: <Satisfied?No Consultant. Media:Labor union;Public IYes Operation Finish retrofit and start regular operation Fig.1.The phases and involved stakeholders in green retrofit projects. based model originally simulated the dynamic opinions of the through the literature review and structured interviews with ex- stakeholders that were influenced by their interactions to allow the perts who are experienced in green retrofit projects.The incentives, adjustment of their values on three dimensions (i.e.,cost,energy, which can be defined as the potential profits of green retrofit,can and comfort)to an optimal retrofit decision.Fuerst and McAllister be classified into three categories,namely,1)direct short-term (2011)analyzed the rent,cost,and price of the buildings that incentives,2)direct long-term incentives,and 3)indirect in- were influenced by green retrofit and tried to define an appropriate centives.The first category refers to the incentives related to the compensation to satisfy both owners and occupiers. economic benefits that can be reaped in a short time,such as high rent,low maintenance cost,and tax reduction.The second category 2.2.Incentives and drivers for the owners and occupiers also refers to economic benefits,but these ones are gained in the long term (i.e..high occupancy rate,asset value raise,and It is important to understand the specific incentives and barriers longevity).The last category includes other incentives related to the for the owners and occupiers as they may have varying and con- social and environmental influence rather than the economic in- flicting opinions on whether a building should be retrofitted and terests.The varying incentives for the owners and occupiers are illustrated in Table 1. when and how the retrofit should be implemented.The owners may be motivated to implement retrofit projects by high rent and occupancy rate (Fuerst and McAllister,2011:Thomas,2010),tax 2.3.Barriers and resistances for the owners and occupiers reduction (Fuerst and McAllister.2011),and reputation enhance- ment (Gucyeter and Gunaydin,2012).Alternatively,the occupiers Other than the aforementioned split incentives,some barriers may be interested in energy cost saving(Ma et al..2012:Newsham (i.e.,cost or potential resistances)exist,and they adversely affect et al.,2009:Rey.2004),low rent(Menassa and Baer.2014),and the decision making in green retrofit projects.Similar to incentives, productivity (Fuerst and McAllister,2011:Thomas,2010;Xu et al.. the main barriers for the owners and occupiers are identified 2011).Conflicts are probably raised when the owners invest in through the literature review and structured interviews.These energy efficiency retrofit,but most direct benefits of energy cost barriers are classified into two categories,namely,the 1)direct/ saving are received by the occupiers. economic barriers,and 2)indirect barriers.The first category de- The main incentives for owners and occupiers are identified notes barriers directly related to the economic problems,such as

based model originally simulated the dynamic opinions of the stakeholders that were influenced by their interactions to allow the adjustment of their values on three dimensions (i.e., cost, energy, and comfort) to an optimal retrofit decision. Fuerst and McAllister (2011) analyzed the rent, cost, and price of the buildings that were influenced by green retrofit and tried to define an appropriate compensation to satisfy both owners and occupiers. 2.2. Incentives and drivers for the owners and occupiers It is important to understand the specific incentives and barriers for the owners and occupiers as they may have varying and con- flicting opinions on whether a building should be retrofitted and when and how the retrofit should be implemented. The owners may be motivated to implement retrofit projects by high rent and occupancy rate (Fuerst and McAllister, 2011; Thomas, 2010), tax reduction (Fuerst and McAllister, 2011), and reputation enhance￾ment (Gucyeter and Gunaydin, 2012). Alternatively, the occupiers may be interested in energy cost saving (Ma et al., 2012; Newsham et al., 2009; Rey, 2004), low rent (Menassa and Baer, 2014), and productivity (Fuerst and McAllister, 2011; Thomas, 2010; Xu et al., 2011). Conflicts are probably raised when the owners invest in energy efficiency retrofit, but most direct benefits of energy cost saving are received by the occupiers. The main incentives for owners and occupiers are identified through the literature review and structured interviews with ex￾perts who are experienced in green retrofit projects. The incentives, which can be defined as the potential profits of green retrofit, can be classified into three categories, namely, 1) direct short-term incentives, 2) direct long-term incentives, and 3) indirect in￾centives. The first category refers to the incentives related to the economic benefits that can be reaped in a short time, such as high rent, low maintenance cost, and tax reduction. The second category also refers to economic benefits, but these ones are gained in the long term (i.e., high occupancy rate, asset value raise, and longevity). The last category includes other incentives related to the social and environmental influence rather than the economic in￾terests. The varying incentives for the owners and occupiers are illustrated in Table 1. 2.3. Barriers and resistances for the owners and occupiers Other than the aforementioned split incentives, some barriers (i.e., cost or potential resistances) exist, and they adversely affect the decision making in green retrofit projects. Similar to incentives, the main barriers for the owners and occupiers are identified through the literature review and structured interviews. These barriers are classified into two categories, namely, the 1) direct/ economic barriers, and 2) indirect barriers. The first category de￾notes barriers directly related to the economic problems, such as Fig. 1. The phases and involved stakeholders in green retrofit projects. 1302 X. Liang et al. / Journal of Cleaner Production 137 (2016) 1300e1312

X.Liang et aL Journal of Cleaner Production 137(2016)1300-1312 1303 Table 1 The incentives of owners and occupiers in green retrofit Stakeholders Direct short-term incentives Direct long-term incentives Indirect incentives Owners Dow1:Higher rent Dow5:Higher occupancy rate Dow9:Reputation enhancement (Fuerst and McAllister,2011:Thomas,2010) (Fuerst and McAllister.2011: (Gucveter and Gunavdin 2012) Dow2:Lower maintenance cost Ma et al,2012:Thomas.2010) Dow10:Social responsibility (Alanne.2004:Lapinski et al.2006: Dow6:Risks reduction (Davies and Osmani,2011) Ouyang et al,2011:Rey.2004) (avoid premature obsolescence,energy cost increasing) Dow11:Occupiers satisfaction Dow3:Subsidies/tax reduction (Fuerst and McAllister,2011) (Thomas,2010:Xu et al.,2011) (Fuerst and McAllister,2011:Ouyang et al.,2011) Dow7:Longevity (Kaklauskas et aL,2004: Dow4:Return on investment (ROl) Mickaityte et al,2008) (Entrop et al,2010:Kaklauskas et al,2004: Dow8:Asset value raise (Miller and Buys,2008) Miller and Buys,2008) Occupiers Doc1:Lower total cost Doc2:Productivity improvement Doc3:Comfort enhancement (including potentially higher rent (Fuerst and McAllister.2011:Lapinski et al,2006: (Wang et aL,2010:Xu et aL,2011) and lower energy cost) Thomas,2010:Xu et al,2011) Doc4:Social responsibility (Caccavelli and Gugerli,2002: (Davies and Osmani,2011) Juan et al,2010:Newsham et al..2009:Rey,2004) high retrofit cost,finite capital,and long payback periods.The buildings,the occupiers are the main stakeholders in the green second category refers to the barriers indirectly related to economy. retrofit for existing buildings,which can be classified into three such as lack of building information,and highly complex design categories according to occupancy type,namely,owner-occupied. analysis and solution.The specific barriers for the owners and oc- single-occupied (not by owner).and multi-occupied (Rhoads. cupiers are shown in Table 2. 2010).These occupancy types can influence the position and po- wer of the stakeholders,benefit distribution,transaction cost,and 2.4.Summary of the review other factors in decision making.In the case of owner-occupied buildings the owners can make decisions completely by them- The abovementioned studies (Fuerst and McAllister,2011: selves.The single-occupied buildings are rented to a single tenant. Stephan and Menassa,2013,2014)are helpful in understanding probably a company or institute in a relatively large scale.The the important roles of the owners and occupiers and their varying single tenants significantly influence the profit of the building interests in green retrofit.However,understanding alone is not owners and therefore have relatively high negotiation capabilities. sufficient to reveal the underlying logic of the industry's reluctance A retrofit decision must be agreed to by the single occupier. otherwise,the decision becomes difficult to implement.The multi- to conduct green retrofit.The shortcomings of the existing studies are explained in the succeeding paragraphs occupied buildings are occupied by numerous tenants who are First,the existing studies generally overlooked the role of the commonly in a small scale.Every occupier only rents a small part of owners and occupiers in green retrofit at the initial phase.Many the building,and the rent from an individual occupier is not suffi- green retrofit plans were canceled at the beginning stage because of cient to influence the profit of owners.In this case,the owner controversies among the owners and occupiers.Therefore,the dominates the decision-making process,as such,if small occupiers behaviors of the owners and occupiers at the initial phase must be do not agree with the owner,they can only choose to "vote with their feet,"that is,to terminate the contract and move out.Based on re-examined to clearly identify the underlying reasons.Focusing on one phase of green retrofit can also help decompose and simplify the above analysis,the influences of occupancy types on the green the decision-making problems as well as achieve relatively accurate retrofit decisions at the initial phase must not be overlooked. Finally,existing studies mainly focused on the effects of cost. results. Second,the existing studies disregarded the complicated be- energy saving,and comfort on green retrofit(Stephan and Menassa. haviors of the owners and occupiers under different interaction 2014).However,some other factors likewise play important roles in relationships,or different occupancy types.Unlike the case of new green retrofit decisions,such as reputation enhancement,risk Table 2 The barriers of owners and occupiers in green retrofit Stakeholders Direct/Economic barriers Indirect barriers Owners Bow1:High retrofit cost Bow6:Highly complex design (Lapinski et al.2006;Menassa,2011:Xu et al,2011) analysis and solution(Davies and Osmani,2011: Bow2:Long payback periods Kasivisvanathan et al.,2012:Lapinski et al.,2006) (Kasivisvanathan et al.,2012:Menassa.2011) Bow7:Lack of building information Bow3:Finite capital (Davies and Osmani,2011: (Davies and osmani 2011 Kasivisvanathan et aL,2012;Menassa,2011: Kasivisvanathan et al.,2012:Menassa,2011) Stiess and Dunkelberg.2013) Bow8:Lack of retrofit experience Bow4:Interruptions in operations (Ali et al,2008:Korkmaz et al,2010) (Kasivisvanathan et al,2012:Miller and Buys,2008) Bow5:Risk of retrofits (Menassa,2011) Occupiers Boc1:Higher rent (Fuerst and McAllister,2011:Thomas,2010) Boc4:Lack of understanding or interest Boc2:Interruptions in operations(Kasivisvanathan et al..2012: about environment(Davies and Osmani.2011) Miller and Buys,2008) Boc5:Lack of information (Davies and Osmani,2011: Boc3:Risk of retrofits Kasivisvanathan et al,2012:Menassa,2011) (energy may not be saved by retrofit)(Menassa.2011) Boc6:Possibility of relocation (occupiers may not bear interruptions or higher rent) (Fuerst and McAllister.2011)

high retrofit cost, finite capital, and long payback periods. The second category refers to the barriers indirectly related to economy, such as lack of building information, and highly complex design analysis and solution. The specific barriers for the owners and oc￾cupiers are shown in Table 2. 2.4. Summary of the review The abovementioned studies (Fuerst and McAllister, 2011; Stephan and Menassa, 2013, 2014) are helpful in understanding the important roles of the owners and occupiers and their varying interests in green retrofit. However, understanding alone is not sufficient to reveal the underlying logic of the industry's reluctance to conduct green retrofit. The shortcomings of the existing studies are explained in the succeeding paragraphs. First, the existing studies generally overlooked the role of the owners and occupiers in green retrofit at the initial phase. Many green retrofit plans were canceled at the beginning stage because of controversies among the owners and occupiers. Therefore, the behaviors of the owners and occupiers at the initial phase must be re-examined to clearly identify the underlying reasons. Focusing on one phase of green retrofit can also help decompose and simplify the decision-making problems as well as achieve relatively accurate results. Second, the existing studies disregarded the complicated be￾haviors of the owners and occupiers under different interaction relationships, or different occupancy types. Unlike the case of new buildings, the occupiers are the main stakeholders in the green retrofit for existing buildings, which can be classified into three categories according to occupancy type, namely, owner-occupied, single-occupied (not by owner), and multi-occupied (Rhoads, 2010). These occupancy types can influence the position and po￾wer of the stakeholders, benefit distribution, transaction cost, and other factors in decision making. In the case of owner-occupied buildings the owners can make decisions completely by them￾selves. The single-occupied buildings are rented to a single tenant, probably a company or institute in a relatively large scale. The single tenants significantly influence the profit of the building owners and therefore have relatively high negotiation capabilities. A retrofit decision must be agreed to by the single occupier, otherwise, the decision becomes difficult to implement. The multi￾occupied buildings are occupied by numerous tenants who are commonly in a small scale. Every occupier only rents a small part of the building, and the rent from an individual occupier is not suffi- cient to influence the profit of owners. In this case, the owner dominates the decision-making process, as such, if small occupiers do not agree with the owner, they can only choose to “vote with their feet,” that is, to terminate the contract and move out. Based on the above analysis, the influences of occupancy types on the green retrofit decisions at the initial phase must not be overlooked. Finally, existing studies mainly focused on the effects of cost, energy saving, and comfort on green retrofit (Stephan and Menassa, 2014). However, some other factors likewise play important roles in green retrofit decisions, such as reputation enhancement, risk Table 1 The incentives of owners and occupiers in green retrofit. Stakeholders Direct short-term incentives Direct long-term incentives Indirect incentives Owners DOW1: Higher rent (Fuerst and McAllister, 2011; Thomas, 2010) DOW2: Lower maintenance cost (Alanne, 2004; Lapinski et al., 2006; Ouyang et al., 2011; Rey, 2004) DOW3: Subsidies/tax reduction (Fuerst and McAllister, 2011; Ouyang et al., 2011) DOW4: Return on investment (ROI) (Entrop et al., 2010; Kaklauskas et al., 2004; Miller and Buys, 2008) DOW5: Higher occupancy rate (Fuerst and McAllister, 2011; Ma et al., 2012; Thomas, 2010) DOW6: Risks reduction (avoid premature obsolescence, energy cost increasing) (Fuerst and McAllister, 2011) DOW7: Longevity (Kaklauskas et al., 2004; Mickaityte et al., 2008) DOW8: Asset value raise (Miller and Buys, 2008) DOW9: Reputation enhancement (Gucyeter and Gunaydin, 2012) DOW10: Social responsibility (Davies and Osmani, 2011) DOW11: Occupiers satisfaction (Thomas, 2010; Xu et al., 2011) Occupiers DOC1: Lower total cost (including potentially higher rent and lower energy cost) (Caccavelli and Gugerli, 2002; Juan et al., 2010; Newsham et al., 2009; Rey, 2004) DOC2: Productivity improvement (Fuerst and McAllister, 2011; Lapinski et al., 2006; Thomas, 2010; Xu et al., 2011) DOC3: Comfort enhancement (Wang et al., 2010; Xu et al., 2011) DOC4: Social responsibility (Davies and Osmani, 2011) Table 2 The barriers of owners and occupiers in green retrofit. Stakeholders Direct/Economic barriers Indirect barriers Owners BOW1: High retrofit cost (Lapinski et al., 2006; Menassa, 2011; Xu et al., 2011) BOW2: Long payback periods (Kasivisvanathan et al., 2012; Menassa, 2011) BOW3: Finite capital (Davies and Osmani, 2011; Kasivisvanathan et al., 2012; Menassa, 2011; Stiess and Dunkelberg, 2013) BOW4: Interruptions in operations (Kasivisvanathan et al., 2012; Miller and Buys, 2008) BOW5: Risk of retrofits (Menassa, 2011) BOW6: Highly complex design analysis and solution (Davies and Osmani, 2011; Kasivisvanathan et al., 2012; Lapinski et al., 2006) BOW7: Lack of building information (Davies and Osmani, 2011; Kasivisvanathan et al., 2012; Menassa, 2011) BOW8: Lack of retrofit experience (Ali et al., 2008; Korkmaz et al., 2010) Occupiers BOC1: Higher rent (Fuerst and McAllister, 2011; Thomas, 2010) BOC2: Interruptions in operations (Kasivisvanathan et al., 2012; Miller and Buys, 2008) BOC3: Risk of retrofits (energy may not be saved by retrofit) (Menassa, 2011) BOC4: Lack of understanding or interest about environment (Davies and Osmani, 2011) BOC5: Lack of information (Davies and Osmani, 2011; Kasivisvanathan et al., 2012; Menassa, 2011) BOC6: Possibility of relocation (occupiers may not bear interruptions or higher rent) (Fuerst and McAllister, 2011) X. Liang et al. / Journal of Cleaner Production 137 (2016) 1300e1312 1303

1304 X.Liang et aL Journal of Cleaner Production 137 (2016)1300-1312 reduction,and transaction costs.In particular,some large-scale motivation and engagement of people even under incentive pol- companies attach significant importance to reputation and risk. icies.To determine the reasons underlying this problem,this study According to our interview,public influence and reputation limits the research scope to the owners and occupiers at the initial enhancement are assigned the highest priority in the green retrofit phase.The behaviors of these decision makers are specified in project of China Recourse Headquarter in Hong Kong,which was Sections 1 and 2. completed in 2012. Second,this study analyzes the main possible influential factors The abovementioned limitations present difficulties in under- of green retrofit decisions by conducting literature review and standing why the existing identified reasons would result in the expert interview.A series of semi-structured interviews are con- reluctance to implement green retrofit in the industry.To mitigate ducted with 19 experts from Hong Kong and Mainland China,who such deficiency.this study aims to reveal the underlying logic of have participated in green retrofit projects as project manager, industry's reluctance to conduct green retrofit by analyzing the designer,facility manager,contract manager,contractor,and third- behaviors of owners and occupiers at the initial phase.This study party consultant authorized by the government to audit projects. differs from the previous ones in that it explores the behaviors of The analysis results are discussed in Section 2. the direct decision makers under current market constraints Third,the identified problem in reality is mapped to a theoret- through game analysis,rather than identifying the willingness of ical model by game theory.Most owners and occupiers prefer to the stakeholders or retrofit-related problems through survey.The show their positive attitudes rather than reluctance towards green factors affecting the behaviors of these decision makers,including retrofit to the public.This condition explains why the major cases in economic,environmental,and social factors,in the short and long the existing studies are successful retrofit projects.Constrained by terms are systemically analyzed and selected based on the litera- such realities,we adopt game theory to investigate the research ture review.Game theory is used to model the behaviors of the question using theoretical model and logical deduction rather than owners and occupiers under different occupancy types.The results an empirical study.The data for the game analysis are primarily of this study are critical to the analysis of the feasibility of a building obtained from the existing studies and interviews about real cases. retrofit and may serve as a foundation for formulating more energy This process is useful in overcoming the shortages of limited data as efficiency strategies and policies. a relative value rather than absolute value that is needed in the game analysis.The probable decisions of the owners and occupiers 3.Methodology under the three occupancy types are discussed in Sections 4-6. Finally,the result interpretations and uncertainty discussion are 3.1.Research design presented in Section 7.The conclusions and limitations of this study are emphasized in Section 8. To fulfill the research aim,this study is conducted with four procedures,which are demonstrated in Fig.2. 3.2.Game theory First,this study identifies the research problem and research gap through the literature review.The relevant literature asserts Game theory is used to model the decision behaviors of owners that,a key problem in promoting green retrofit is the low and occupiers in green retrofit.Previous studies have determined Steps Methods/Tools Outcomes Step 1 Identifying the Research problem: research problem and Literature review Research objectives; research gap Key stakeholders Step 2 Analyzing the impact Expert interview: Occupancy types; factors Literature review Incentives and divers Barriers and resistances Step 3 Mapping the real Nash Equilibrium under problem to theoretical Game theory different occupancy model types Step 4 Discussing the Inductive Result interpretations analysis results interpretation and suggestions; Future directions Fig.2.Schematic of research process

reduction, and transaction costs. In particular, some large-scale companies attach significant importance to reputation and risk. According to our interview, public influence and reputation enhancement are assigned the highest priority in the green retrofit project of China Recourse Headquarter in Hong Kong, which was completed in 2012. The abovementioned limitations present difficulties in under￾standing why the existing identified reasons would result in the reluctance to implement green retrofit in the industry. To mitigate such deficiency, this study aims to reveal the underlying logic of industry's reluctance to conduct green retrofit by analyzing the behaviors of owners and occupiers at the initial phase. This study differs from the previous ones in that it explores the behaviors of the direct decision makers under current market constraints through game analysis, rather than identifying the willingness of the stakeholders or retrofit-related problems through survey. The factors affecting the behaviors of these decision makers, including economic, environmental, and social factors, in the short and long terms are systemically analyzed and selected based on the litera￾ture review. Game theory is used to model the behaviors of the owners and occupiers under different occupancy types. The results of this study are critical to the analysis of the feasibility of a building retrofit and may serve as a foundation for formulating more energy efficiency strategies and policies. 3. Methodology 3.1. Research design To fulfill the research aim, this study is conducted with four procedures, which are demonstrated in Fig. 2. First, this study identifies the research problem and research gap through the literature review. The relevant literature asserts that, a key problem in promoting green retrofit is the low motivation and engagement of people even under incentive pol￾icies. To determine the reasons underlying this problem, this study limits the research scope to the owners and occupiers at the initial phase. The behaviors of these decision makers are specified in Sections 1 and 2. Second, this study analyzes the main possible influential factors of green retrofit decisions by conducting literature review and expert interview. A series of semi-structured interviews are con￾ducted with 19 experts from Hong Kong and Mainland China, who have participated in green retrofit projects as project manager, designer, facility manager, contract manager, contractor, and third￾party consultant authorized by the government to audit projects. The analysis results are discussed in Section 2. Third, the identified problem in reality is mapped to a theoret￾ical model by game theory. Most owners and occupiers prefer to show their positive attitudes rather than reluctance towards green retrofit to the public. This condition explains why the major cases in the existing studies are successful retrofit projects. Constrained by such realities, we adopt game theory to investigate the research question using theoretical model and logical deduction rather than an empirical study. The data for the game analysis are primarily obtained from the existing studies and interviews about real cases. This process is useful in overcoming the shortages of limited data as a relative value rather than absolute value that is needed in the game analysis. The probable decisions of the owners and occupiers under the three occupancy types are discussed in Sections 4e6. Finally, the result interpretations and uncertainty discussion are presented in Section 7. The conclusions and limitations of this study are emphasized in Section 8. 3.2. Game theory Game theory is used to model the decision behaviors of owners and occupiers in green retrofit. Previous studies have determined Methods/Tools Research problem; Research objectives; Key stakeholders Steps Identifying the research problem and research gap Analyzing the impact factors Mapping the real problem to theoretical model Discussing the analysis results Literature review Expert interview; Literature review Game theory Inductive interpretation Outcomes Occupancy types; Incentives and divers; Barriers and resistances Nash Equilibrium under different occupancy types Result interpretations and suggestions; Future directions Step 1 Step 2 Step 3 Step 4 Fig. 2. Schematic of research process. 1304 X. Liang et al. / Journal of Cleaner Production 137 (2016) 1300e1312

X.Liang et aL Journal of Cleaner Production 137 (2016)1300-1312 1305 that many owners and occupiers intend to implement green 1)Players:Two players are involved in the game,namely,the retrofit,but only a few can reach a consensus and continue its owner and the occupier,as elaborated in Section 2.1. implementation.Although other methods can be used to identify 2)Strategies:In general,the owner is the entity who establishes the reasons behind the actions,game theory focuses on players the initiative to retrofit and provides the initial retrofit plans. different actions influenced by the actions of other players.In the However,the occupiers have become increasingly active in beginning of green retrofit,owners and occupiers,as key decision green retrofit projects in recent years.Given that numerous makers,are interdependent and have different interests on the buildings have been retrofitted over the past few years,the issue.For example,when the owners intend to administer green successful experiences have attracted the occupiers because of retrofit to improve their social reputation or to reduce maintenance the learning effect.In 2015,81 occupiers actively raised 16 cost,they have to consider the rent contracts with occupiers.On the million CNY to finish the green retrofit project in the Interna- contrary,when the occupiers want to implement green retrofit to tional Trade Center of Shenzhen,China(Xiao,2015).Therefore, save energy cost,they have to consider the attitudes of the owners. both the owner and occupier in this game have two strategies, Thus,the final green retrofit decision depends on the actions of namely,"initiative to retrofit"and "reluctant to retrofit"The both the owners and occupiers.In this case,game theory can be former strategy is initiative to conduct green retrofit,whereas adopted to analyze the strategies of both parties.The results of the the latter is the resistance to the implementation of green game theory analysis presented in this paper can provide causal retrofit but keeps regular operation. and solid proof for green retrofit decisions. 3)Payoffs:The payoffs of owners and occupiers depend on their Game theory was established to identify the optimal solutions respective strategies,which are shown in the payoff matrix in for economic behaviors (Von Neumann and Morgenstern.1945). Fig.3.Bow11.Bow12.Bow21 and Bow22 represent the benefits of an Nash (1950)developed a definition of an "optimum"strategy for owner under different strategies, whereas multiplayer games.This strategy,which is well-known as the "Nash Boc11.Boc12.Boc21.and Boc22 represent the benefits of an occupier. Equilibrium",indicates that every player cannot obtain a benefit by changing his or her own action:thus,the equilibrium is stable (Healy.2006).Nash equilibrium,which is a type of game theory,is 4.Retrofit decision for owner-occupied building generally used to analyze the competition or collaboration prob- lems between two decision makers,such as the prisoner's dilemma The owner-occupied building is first discussed,because it is the (Fudenberg and Tirole,1991).Three basic elements exist in game simplest occupancy type and can be used as a baseline.In this type theory (ie.,player,strategy,and payoff).A player,who assumes of building,the occupiers also own the building,and they can make absolutely rational self-interest,is an individual participant in the retrofit decisions by themselves without negotiation.The income of decision-making of strategic choices.Strategy is the choice or ac- the owners comes from the energy cost saving.maintenance cost tion of a player,that can either be a pure or mixed strategy in saving.building value increase,and public impact.The costs are the certain probabilities.Payoff is the interest that a player accrues by retrofit investment and operation disturbance.The rent,occupancy adopting a strategy (Peng et al.,2014).Payoffs,which are quanti- rate,and turnover rate are not considered in this owner-occupied tative,are normally described by a payoff matrix to illustrate the situation.These variables are described in Table 3.The column of interest of a player based on all decisions. "Driver/Barrier"shows the corresponding relation between the Game theory is widely used in research related to sustainable variables and the drivers or barriers illustrated in Tables 1 and 2 development and green building,particularly with regard to the This study assumes that the decision to retrofit depends on the relationship among the stakeholders and their decision making.Gu benefits that the decision makers can reap from retrofit.Therefore, et al.(2009)analyzed the strategies for energy-efficient housing non-economic factors,such as the lack of building information developments with game theory by integrating four players. (Bow6)and lack of retrofit experience(Bow7).are not considered in namely,the administration,the developers,the architects,and the this analysis.The subsidies and tax reduction (Dow3)are not inhabitants.The study identified several crucial issues in energy- included in the variables,because the model focuses on the retrofit efficient building development and indicated that achieving the decisions without market interventions.In fact,the Chinese gov- energy efficiency objective is difficult if the actions of all the players ernment only provides incentive funds for the energy-efficiency are based on their respective rational self-interest(Gu et al.,2009). retrofit of the residential buildings in the northern heating area However,this conclusion is relatively general and is not based on a of China(Zhou et al.,2010)and of the public buildings in a few pilot concrete analysis of relationships and interests.Some studies have cities.Most areas are not funded by government,a condition that is specified and quantified the interest of the players in games (Li consistent with this model.The risks of retrofits(Dow5/Bow6)are et al 2011:Mohsenian-Rad et al..2010).Li et al.(2011)proposed likewise not considered in this analysis,which are discussed in a game theory model to analyze the energy-saving building market Section 7.3. in China through a game of customers and developers,whose in- Bocan then be calculated with the following formula,where terests were calculated quantitatively.Game theory has also been the superscript "oo"represents the owner-occupied condition: used to evaluate and simulate energy consumption (Mohsenian- B80=S8°+S8°+AVo0+Ap0-10-D0 (1) Rad et al.,2010:Soliman and Leon-Garcia,2014).However,the existing studies based on game theory are primarily for new energy All variables represent the life cycle value,which is an efficient efficiency building rather than for green retrofit.To elaborate the method in conducting an economic analysis of the building retrofit use of game theory in this study,the research problem definition issues (Ouyang et al,2011).Various factors may influence the de- are presented in the following sub-section. cision making of retrofit in different levels.The factors related to the direct economical profit.such as Doo and seo,may be 3.3.Problem definition considered as the high priority;otherwise,they are regarded as the relatively low priority (e.g..AP0).The difference among these The problem of developing a strategy on green retrofit is defined factors depends on specific projects and on the evaluation of the as a non-cooperative game between the owners and occupiers of owners.When Bo>0 in this owner-occupied scenario,the owners the existing buildings.The critical elements for the game analysis can benefit from the green retrofit and will choose it as the strategy. are specified below. Otherwise,the green retrofit will not be implemented.The building

that many owners and occupiers intend to implement green retrofit, but only a few can reach a consensus and continue its implementation. Although other methods can be used to identify the reasons behind the actions, game theory focuses on players' different actions influenced by the actions of other players. In the beginning of green retrofit, owners and occupiers, as key decision makers, are interdependent and have different interests on the issue. For example, when the owners intend to administer green retrofit to improve their social reputation or to reduce maintenance cost, they have to consider the rent contracts with occupiers. On the contrary, when the occupiers want to implement green retrofit to save energy cost, they have to consider the attitudes of the owners. Thus, the final green retrofit decision depends on the actions of both the owners and occupiers. In this case, game theory can be adopted to analyze the strategies of both parties. The results of the game theory analysis presented in this paper can provide causal and solid proof for green retrofit decisions. Game theory was established to identify the optimal solutions for economic behaviors (Von Neumann and Morgenstern, 1945). Nash (1950) developed a definition of an “optimum” strategy for multiplayer games. This strategy, which is well-known as the “Nash Equilibrium”, indicates that every player cannot obtain a benefit by changing his or her own action; thus, the equilibrium is stable (Healy, 2006). Nash equilibrium, which is a type of game theory, is generally used to analyze the competition or collaboration prob￾lems between two decision makers, such as the prisoner's dilemma (Fudenberg and Tirole, 1991). Three basic elements exist in game theory (i.e., player, strategy, and payoff). A player, who assumes absolutely rational self-interest, is an individual participant in the decision-making of strategic choices. Strategy is the choice or ac￾tion of a player, that can either be a pure or mixed strategy in certain probabilities. Payoff is the interest that a player accrues by adopting a strategy (Peng et al., 2014). Payoffs, which are quanti￾tative, are normally described by a payoff matrix to illustrate the interest of a player based on all decisions. Game theory is widely used in research related to sustainable development and green building, particularly with regard to the relationship among the stakeholders and their decision making. Gu et al. (2009) analyzed the strategies for energy-efficient housing developments with game theory by integrating four players, namely, the administration, the developers, the architects, and the inhabitants. The study identified several crucial issues in energy￾efficient building development and indicated that achieving the energy efficiency objective is difficult if the actions of all the players are based on their respective rational self-interest (Gu et al., 2009). However, this conclusion is relatively general and is not based on a concrete analysis of relationships and interests. Some studies have specified and quantified the interest of the players in games (Li et al., 2011; Mohsenian-Rad et al., 2010). Li et al. (2011) proposed a game theory model to analyze the energy-saving building market in China through a game of customers and developers, whose in￾terests were calculated quantitatively. Game theory has also been used to evaluate and simulate energy consumption (Mohsenian￾Rad et al., 2010; Soliman and Leon-Garcia, 2014). However, the existing studies based on game theory are primarily for new energy efficiency building rather than for green retrofit. To elaborate the use of game theory in this study, the research problem definition are presented in the following sub-section. 3.3. Problem definition The problem of developing a strategy on green retrofit is defined as a non-cooperative game between the owners and occupiers of the existing buildings. The critical elements for the game analysis are specified below. 1) Players: Two players are involved in the game, namely, the owner and the occupier, as elaborated in Section 2.1. 2) Strategies: In general, the owner is the entity who establishes the initiative to retrofit and provides the initial retrofit plans. However, the occupiers have become increasingly active in green retrofit projects in recent years. Given that numerous buildings have been retrofitted over the past few years, the successful experiences have attracted the occupiers because of the learning effect. In 2015, 81 occupiers actively raised 16 million CNY to finish the green retrofit project in the Interna￾tional Trade Center of Shenzhen, China (Xiao, 2015). Therefore, both the owner and occupier in this game have two strategies, namely, “initiative to retrofit” and “reluctant to retrofit.” The former strategy is initiative to conduct green retrofit, whereas the latter is the resistance to the implementation of green retrofit but keeps regular operation. 3) Payoffs: The payoffs of owners and occupiers depend on their respective strategies, which are shown in the payoff matrix in Fig. 3. Bow11, Bow12, Bow21 and Bow22 represent the benefits of an owner under different strategies, whereas Boc11, Boc12, Boc21, and Boc22 represent the benefits of an occupier. 4. Retrofit decision for owner-occupied building The owner-occupied building is first discussed, because it is the simplest occupancy type and can be used as a baseline. In this type of building, the occupiers also own the building, and they can make retrofit decisions by themselves without negotiation. The income of the owners comes from the energy cost saving, maintenance cost saving, building value increase, and public impact. The costs are the retrofit investment and operation disturbance. The rent, occupancy rate, and turnover rate are not considered in this owner-occupied situation. These variables are described in Table 3. The column of “Driver/Barrier” shows the corresponding relation between the variables and the drivers or barriers illustrated in Tables 1 and 2 This study assumes that the decision to retrofit depends on the benefits that the decision makers can reap from retrofit. Therefore, non-economic factors, such as the lack of building information (BOW6) and lack of retrofit experience (BOW7), are not considered in this analysis. The subsidies and tax reduction (DOW3) are not included in the variables, because the model focuses on the retrofit decisions without market interventions. In fact, the Chinese gov￾ernment only provides incentive funds for the energy-efficiency retrofit of the residential buildings in the northern heating area of China (Zhou et al., 2010) and of the public buildings in a few pilot cities. Most areas are not funded by government, a condition that is consistent with this model. The risks of retrofits (DOW5/BOW6) are likewise not considered in this analysis, which are discussed in Section 7.3. Boo ow can then be calculated with the following formula, where the superscript “oo” represents the owner-occupied condition: Boo ow ¼ Soo e þ Soo o þ DVoo þ DPoo I oo Doo (1) All variables represent the life cycle value, which is an efficient method in conducting an economic analysis of the building retrofit issues (Ouyang et al., 2011). Various factors may influence the de￾cision making of retrofit in different levels. The factors related to the direct economical profit, such as I oo, Doo and Soo e , may be considered as the high priority; otherwise, they are regarded as the relatively low priority (e.g., DPoo). The difference among these factors depends on specific projects and on the evaluation of the owners. When Boo ow >0 in this owner-occupied scenario, the owners can benefit from the green retrofit and will choose it as the strategy. Otherwise, the green retrofit will not be implemented. The building X. Liang et al. / Journal of Cleaner Production 137 (2016) 1300e1312 1305

1306 X.Liang et aL Journal of Cleaner Production 137(2016)1300-1312 Occupiers Initiative to retrofit Reluctant to retrofit Initiative to (BB） (Brl2,Bcl2） retrofit Reluctant to retrofit (Bn2.Be21） (Bnm22,B.22） Fig.3.The payoff matrix of retrofit strategy for owner and occupier. Table 3 The variables of decision in owner-occupied building. Stakeholder Variable Definition Driver/Barrier Owner Bow Benefit of a building owner Dow1/4 Energy saving through green retrofit Dow4 S。 Operation saving through green retrofit Dow2 △V Building value increase through green retrofit Dow7/8 △P Public impact through retrofit Dow9/10/11 1 Investment of building retrofit Bow1/2/3 Disturbance of business during retrofit Bow4 owners and occupiers in this scenario have no game,because the owners.Different from the scenario of an owner-occupied building. decision is made only by the owners. the rent and disturbance of the business during the retrofitting is accrued to the occupiers rather than to the owners.These variables 5.Retrofit decision for single-occupied building are described in Table 4.The non-economic factors,such as comfort enhancement (Doc3).lack of understanding or interest about In the scenario of a single-occupied building.two players. environment (Boc4)and lack of information (Boc5).are not considered. namely,the owner and the single occupier,are in the game for green retrofit decision making.The advantages and disadvantages The payoff matrix for the owners and occupiers in a single- of the green retrofit for the owner are the economic aspects and occupied building is shown in Fig.4,where the superscript "so" represents the single-occupied condition. social influences of the process.The economic aspects include the Based on this adapted matrix,the following scenarios are operation cost saving from green retrofit,building value increasing. and investment.The social influences pertain to the reputation of a investigated to identify the Nash Equilibrium under the single- occupied condition. company,enterprise social responsibility,environmental impact, and other factors related to the society.To confirm the preliminary proposal and decide on the implementation,the owner and the 5.1.The action of the occupier when the owner has the initiative to occupier must communicate and negotiate with each other,which may cost time,money.and human resource.In addition to the retrofit factors illustrated in Section 4,the increasing rent and coordination Green retrofit can be implemented smoothly when the owner costs are the influencing factors for owners in this scenario. has the initiative to implement the approach and the occupier is For single occupiers,the advantages and disadvantages of green also interested in it.The occupier should be willing to pay addi- retrofit pertain primarily to the economic aspects and social in- tional rent to the owner to compensate for additional costs of fluences.Economic aspects include energy cost saving from green implementing green retrofit (Fuerst and McAllister.2011).This retrofit,rent increase,and disturbance of business during the ret- rofitting.Social influences are the same as the social factors of the scenario occurs mostly in government and large private organiza- tions,where sustainable development is considered an essential factor in building selection(Miller and Buys,2008).The occupier Table 4 tends to pay additional rent 4RIn cases where the owner de- The variables of decision in single-occupied building. cides to implement green retrofit but the occupier is reluctant, Stakeholder Variable Definition Driver/Barrier green retrofit can still be implemented because the owner is the main decision-maker.The reluctant occupier,as a large organiza- Owner △R Increased rent by retrofit Dow1/5 tion,cannot move out because of the long rental contract and high Ce Coordination cost for retrofit Bow1 Occupier Benefit of building occupier Doc1/2 relocation cost,but could take the position of non-cooperation. Se Energy saving through green retrofit Docl Given the strong influence of the owner's income and profit,the R Increased rent through retrofit Bocl single occupier will have relatively high negotiation capacity and D Disturbance of business during retrofit Boc2 AP will not pay additional rent 4R or other alternative compensations Public impact through retrofit Doc4 to the owner when he/she does not want to cooperate in green

owners and occupiers in this scenario have no game, because the decision is made only by the owners. 5. Retrofit decision for single-occupied building In the scenario of a single-occupied building, two players, namely, the owner and the single occupier, are in the game for green retrofit decision making. The advantages and disadvantages of the green retrofit for the owner are the economic aspects and social influences of the process. The economic aspects include the operation cost saving from green retrofit, building value increasing, and investment. The social influences pertain to the reputation of a company, enterprise social responsibility, environmental impact, and other factors related to the society. To confirm the preliminary proposal and decide on the implementation, the owner and the occupier must communicate and negotiate with each other, which may cost time, money, and human resource. In addition to the factors illustrated in Section 4, the increasing rent and coordination costs are the influencing factors for owners in this scenario. For single occupiers, the advantages and disadvantages of green retrofit pertain primarily to the economic aspects and social in- fluences. Economic aspects include energy cost saving from green retrofit, rent increase, and disturbance of business during the ret￾rofitting. Social influences are the same as the social factors of the owners. Different from the scenario of an owner-occupied building, the rent and disturbance of the business during the retrofitting is accrued to the occupiers rather than to the owners. These variables are described in Table 4. The non-economic factors, such as comfort enhancement (DOC3), lack of understanding or interest about environment (BOC4) and lack of information (BOC5), are not considered. The payoff matrix for the owners and occupiers in a single￾occupied building is shown in Fig. 4, where the superscript “so” represents the single-occupied condition. Based on this adapted matrix, the following scenarios are investigated to identify the Nash Equilibrium under the single￾occupied condition. 5.1. The action of the occupier when the owner has the initiative to retrofit Green retrofit can be implemented smoothly when the owner has the initiative to implement the approach and the occupier is also interested in it. The occupier should be willing to pay addi￾tional rent to the owner to compensate for additional costs of implementing green retrofit (Fuerst and McAllister, 2011). This scenario occurs mostly in government and large private organiza￾tions, where sustainable development is considered an essential factor in building selection (Miller and Buys, 2008). The occupier tends to pay additional rent DRso oc11. In cases where the owner de￾cides to implement green retrofit but the occupier is reluctant, green retrofit can still be implemented because the owner is the main decision-maker. The reluctant occupier, as a large organiza￾tion, cannot move out because of the long rental contract and high relocation cost, but could take the position of non-cooperation. Given the strong influence of the owner's income and profit, the single occupier will have relatively high negotiation capacity and will not pay additional rent DR or other alternative compensations to the owner when he/she does not want to cooperate in green Fig. 3. The payoff matrix of retrofit strategy for owner and occupier. Table 3 The variables of decision in owner-occupied building. Stakeholder Variable Definition Driver/Barrier Owner Bow Benefit of a building owner DOW1/4 Se Energy saving through green retrofit DOW4 So Operation saving through green retrofit DOW2 DV Building value increase through green retrofit DOW7/8 DP Public impact through retrofit DOW9/10/11 I Investment of building retrofit BOW1/2/3 D Disturbance of business during retrofit BOW4 Table 4 The variables of decision in single-occupied building. Stakeholder Variable Definition Driver/Barrier Owner DR Increased rent by retrofit DOW1/5 Cc Coordination cost for retrofit BOW1 Occupier Boc Benefit of building occupier DOC1/2 Se Energy saving through green retrofit DOC1 DR Increased rent through retrofit BOC1 D Disturbance of business during retrofit BOC2 DP Public impact through retrofit DOC4 1306 X. Liang et al. / Journal of Cleaner Production 137 (2016) 1300e1312

X.Liang et aL Journal of Cleaner Production 137(2016)1300-1312 1307 Occupier Initiative to retrofit Reluctant to retrofit B-S+AVARAP--CH Bm2=S2+Ar+△P-Ig-C0 Initiative to retrofit B2-Sm+△P-DT-AR1 Bac2-Sc2+AP2-D2 Reluctant to BSAAP B22=0 retrofit B01=S2,+A-D-I份 BC2=0 Fig.4.The payoff matrix for owner and occupier in single-occupied building. retrofit during the contract period.To sum up,regardless of the the single occupier generally has a long rental contract,which initiative of the occupier to implement green retrofit,green retrofit makes possible for reaping returns on investment to cover the is implemented only if the owner wants to. retrofit cost.However,the costs of retrofit,which are investment The benefit and cost of green retrofit are similar in two condi- and disturbance of business during retrofitDs,are short term and tions,but the difference is whether an increment exists in rent after definite,whereas the profits of retrofit,which are energy cost retrofit.According to previous case studies,rent,price,and occu- saving S.and public impact AP.are long term and uncertain.In pancy rate of a building are positively related to its green feature,as addition.energy saving is not reliable because of the contract shown in Table 5.Therefore,this study assumes that green retrofit period.Rent contract may be terminated in several years,which could raise building value(Miller and Buys.2008),occupancy rate, means the occupier can only obtain S for several years rather and rent(Thomas,2010).The key information can be summarized than for the whole life cycle of a building.Specifically,the duration as follows: of reaping profit may be shorter than payback time.Given that most investors are often reluctant to take challenges(Rhoads,2010),they S1+△P9-D8=S92+AP8-D8,△R011≥0 (2) tend to assign more weight to certain cost in the short term than to Referring these data to the formula in Fig.4.the matrix reveals uncertain profit in the long term.The key information can be that. summarized as follows: S81+△P骝-D9-AR11≤S92+AP8-D2, (3) S21+AP%-D2-≤0，B%222=0 (5) Or namely. B11≤B212 (4) B821≤B822 (6) According to Formula(4),the occupier should choose"reluctant According to Formula(6),the occupier should choose "reluctant to retrofit"or non-operation if the owner has the initiative to to retrofit"when the owner is reluctant to green retrofit because of implement green retrofit. the risk in payback period. Based on Formulas(4)and(6),the best interests of the occupier are served by the "reluctant to retrofit"action regardless of the 5.2.The action of the occupier when the owner is reluctant to action the owner takes,which means "reluctant to retrofit"is the retrofit dominant strategy for the occupier(Myerson,2013).Therefore,the action of the owner is discussed in the following section based on If both the owner and the occupier are reluctant,the building the situation where the occupier is reluctant to retrofit. operates without the innovation.Thus,neither cost nor profit exists according to green retrofit,that is,B=0.In the situation where 5.3.The action of the owner when the occupier is reluctant to the owner is reluctant and does not want to invest,but the occupier retrofit is active in sustainable development,the occupier can choose to invest in the retrofit project.The single occupier is mainly a large As mentioned earlier,if the owner and the occupier are reluctant organization that has high economic strength to support green to green retrofit,then green retrofit is not implemented.and the retrofit (Miller and Buys,2008).In addition to economic strength. owner obtains neither cost nor profit,such that B22=0.If the Table 5 The premium of rent.value and occupancy rate in green buildings. Literature Rental premium Value premium Occupancy rate premium Miller et al.(2008) 9 No premium 2-4% Eichholtz et al.(2010) 33% 19% NA Pivo and Fisher (2010) 2.7% 8.5% NA Wiley et aL.(2010) 7-17% NA 10-18% Fuerst and McAllister (2011) 4-5% 25-26% 1-3%

retrofit during the contract period. To sum up, regardless of the initiative of the occupier to implement green retrofit, green retrofit is implemented only if the owner wants to. The benefit and cost of green retrofit are similar in two condi￾tions, but the difference is whether an increment exists in rent after retrofit. According to previous case studies, rent, price, and occu￾pancy rate of a building are positively related to its green feature, as shown in Table 5. Therefore, this study assumes that green retrofit could raise building value (Miller and Buys, 2008), occupancy rate, and rent (Thomas, 2010). The key information can be summarized as follows: Sso e11 þ DPso 11 Dso 11zSso e12 þ DPso 12 Dso 12; DRso oc11 0 (2) Referring these data to the formula in Fig. 4, the matrix reveals that. Sso e11 þ DPso 11 Dso 11 DRso oc11  Sso e12 þ DPso 12 Dso 12; (3) Or Bso oc11  Bso oc12 (4) According to Formula (4), the occupier should choose “reluctant to retrofit” or non-operation if the owner has the initiative to implement green retrofit. 5.2. The action of the occupier when the owner is reluctant to retrofit If both the owner and the occupier are reluctant, the building operates without the innovation. Thus, neither cost nor profit exists according to green retrofit, that is, Bso oc22 ¼ 0. In the situation where the owner is reluctant and does not want to invest, but the occupier is active in sustainable development, the occupier can choose to invest in the retrofit project. The single occupier is mainly a large organization that has high economic strength to support green retrofit (Miller and Buys, 2008). In addition to economic strength, the single occupier generally has a long rental contract, which makes possible for reaping returns on investment to cover the retrofit cost. However, the costs of retrofit, which are investment I so 21 and disturbance of business during retrofitDso 21, are short term and definite, whereas the profits of retrofit, which are energy cost saving Sso e21, and public impact DPso 21, are long term and uncertain. In addition, energy saving Sso e21 is not reliable because of the contract period. Rent contract may be terminated in several years, which means the occupier can only obtain Sso e21 for several years rather than for the whole life cycle of a building. Specifically, the duration of reaping profit may be shorter than payback time. Given that most investors are often reluctant to take challenges (Rhoads, 2010), they tend to assign more weight to certain cost in the short term than to uncertain profit in the long term. The key information can be summarized as follows: Sso e21 þ DPso 21 Dso 21 I so 21  0; Bso oc22 ¼ 0 (5) namely, Bso oc21  Bso oc22 (6) According to Formula (6), the occupier should choose “reluctant to retrofit” when the owner is reluctant to green retrofit because of the risk in payback period. Based on Formulas (4) and (6), the best interests of the occupier are served by the “reluctant to retrofit” action regardless of the action the owner takes, which means “reluctant to retrofit” is the dominant strategy for the occupier (Myerson, 2013). Therefore, the action of the owner is discussed in the following section based on the situation where the occupier is reluctant to retrofit. 5.3. The action of the owner when the occupier is reluctant to retrofit As mentioned earlier, if the owner and the occupier are reluctant to green retrofit, then green retrofit is not implemented, and the owner obtains neither cost nor profit, such that Bso ow22 ¼ 0. If the Fig. 4. The payoff matrix for owner and occupier in single-occupied building. Table 5 The premium of rent, value and occupancy rate in green buildings. Literature Rental premium Value premium Occupancy rate premium Miller et al. (2008) 9% No premium 2e4% Eichholtz et al. (2010) 3.3% 1.9% NA Pivo and Fisher (2010) 2.7% 8.5% NA Wiley et al. (2010) 7e17% NA 10e18% Fuerst and McAllister (2011) 4e5% 25e26% 1e3% X. Liang et al. / Journal of Cleaner Production 137 (2016) 1300e1312 1307

1308 X.Liang et aL Journal of Cleaner Production 137(2016)1300-1312 occupier is not interested in green retrofit but the owner has the 6.1.The action of the owner when the occupier has the initiative to initiative in the approach,then it can be implemented without the retrofit cooperation of the occupier,as discussed in the previous section. Under this condition,the owner cannot reap certain and direct In a situation where the owner does not want to invest on payback from increased rent AR,but can only obtain uncertain, retrofit,but the occupiers want to,a single occupier does not have long-term,and indirect profit from green retrofit,such as operation enough economic capabilities or influential power to implement cost saving building value increase AVse,and public impact the approach.Therefore,if the active occupiers cannot reach Aps.By contrast,investment and coordination cost Cso are consensus with the reluctant owner and cannot implement retrofit certain,short-term and direct,and are thus considered important by themselves,they can only choose to "vote with their feet,"which by the owner because of the risk-adverse characteristic (Rhoads. raises the turnover cost for the owner C.However,this turnover 2010).The key information can be summarized as follows: cost does not increase significantly because small companies consider cost and location as the dominant factors for building S02+△V9+△P8-9-C2≤0，B222-0 7) selection (Rhoads,2010).If occupiers are interested in green retrofit,they will support it and pay additional rent to the owner to namely. compensate for the additional costs (Fuerst and McAllister,2011). Under this condition,the owner can reap a direct payback from the B8w12≤B8w22 (8) increased rent 4R as compensation for the investment.How- ever,the owner must pay the coordination cost to raise the rent. According to Formula(8),the owner should choose "reluctant to Coordination cost C is proportional to the number of occupiers. retrofit"when the occupier is reluctant to green retrofit. The cost is very high when numerous occupiers,various contracts. The preceding analysis in Formula(8)reveals that,in a single- and different rental periods exist (Rhoads.2010),namely. occupied building.the occupier chooses "reluctant to retrofit" C>>0.The key information can be summarized as follows: regardless of what the owner chooses,and the owner is reluctant to retrofit to guarantee his or her interests.Hence,the Nash Equilib- S9+△Vm+△Rm011+△P--C≤0，-C=0 rium for the owner and the single occupier is "reluctant to retrofit," (9) and "reluctant to retrofit,"respectively.Given that a multi-occupied building is another common occupancy type,a comparative study namely. is discussed in the next section B8R11≤B021 (10) According to Formula (10),the owner should choose "reluctant 6.Retrofit decision for multi-occupied building to retrofit"when the initiative of the occupiers is to engage in green retrofit. Given that each occupier is an independent economic entity in this scenario,the owner must play games with each occupier 6.2.The action of the owner when the occupier is reluctant to individually.The occupiers have similar scales,costs,and benefits, retrofit and thus their decisions should not be significantly different. Therefore,occupiers are considered homogeneous in the game.The If both owner and occupiers are reluctant to green retrofit,then game can be abstracted with two players,namely,owner and ho- it is not implemented.The owner obtains neither cost nor profit, mogeneous occupiers.This study does not focus on the differences of occupiers,and the cooperation and games among occupiers are such that B0.When the occupier is not active in green not included as well. retrofit but the owner is,the approach can be implemented without the cooperation of some reluctant occupiers.If some occupiers Different from single-occupied condition,numerous occupiers cannot reach consensus with the owner,they will move out,which exist in a multi-occupied building.In this condition,occupancy is relatively small scale and has a high turnover rate.Therefore,the results in turnover cost C for the owner.Generally,occupancy owner is dominant in the relationship.Occupiers are powerless in rate and rent per unit increase after green retrofit(Ma et al..2012: Thomas,2010).Thus,even though some occupiers move out negotiation and can only choose to "vote with their feet,"which means relocating to another building.Relocation raises additional because of retrofit,the owner can reap increased rentR from the new occupiers.However,high occupancy rate in the future and costs for occupiers and owner in terms of relocation cost and high rent from a new occupier is an indirect and long-term benefit turnover cost respectively.Table 6 describes these additional costs. for the owner.Other profits of retrofit,including operation cost Other advantages and disadvantages of green retrofit are similar to those of the condition of single-occupied building.The payoff saving S,increased building value AV2,and public impact AP are long term and uncertain.By contrast,the costs of retrofit are matrix for owners and occupiers in multi-occupied buildings is short term and definite.The risk-adverse investor does not want to shown in Fig.5,where the superscript "mo"represents the multi- take challenges to compare uncertain interests with specific costs occupied condition: The key information can be summarized as follows: Based on this adapted matrix,the following scenarios can be investigated to identify the Nash Equilibrium under the multi- S88+△V+△R912+△P--C8-C≤0，B922 occupied condition. =0 (11) Table 6 The variables of decision in multi-occupied building. namely, Stakeholder Variable Definition Driver/Barrier Owner Turnover cost of owner Bow1/4 B2≤B22 (12) Occupier Relocation cost of occupier Boc6 According to Formula(12),the owner should choose "reluctant

occupier is not interested in green retrofit but the owner has the initiative in the approach, then it can be implemented without the cooperation of the occupier, as discussed in the previous section. Under this condition, the owner cannot reap certain and direct payback from increased rent DR, but can only obtain uncertain, long-term, and indirect profit from green retrofit, such as operation cost saving Sso o12, building value increase DVso 12, and public impact DPso 12. By contrast, investment Iso 12 and coordination cost Cso c12 are certain, short-term and direct, and are thus considered important by the owner because of the risk-adverse characteristic (Rhoads, 2010). The key information can be summarized as follows: Sso o12 þ DVso 12 þ DPso 12 I so 12 Cso c12  0; Bso oc22 ¼ 0 (7) namely, Bso ow12  Bso ow22 (8) According to Formula (8), the owner should choose “reluctant to retrofit” when the occupier is reluctant to green retrofit. The preceding analysis in Formula (8) reveals that, in a single￾occupied building, the occupier chooses “reluctant to retrofit” regardless of what the owner chooses, and the owner is reluctant to retrofit to guarantee his or her interests. Hence, the Nash Equilib￾rium for the owner and the single occupier is “reluctant to retrofit,” and “reluctant to retrofit,” respectively. Given that a multi-occupied building is another common occupancy type, a comparative study is discussed in the next section. 6. Retrofit decision for multi-occupied building Given that each occupier is an independent economic entity in this scenario, the owner must play games with each occupier individually. The occupiers have similar scales, costs, and benefits, and thus their decisions should not be significantly different. Therefore, occupiers are considered homogeneous in the game. The game can be abstracted with two players, namely, owner and ho￾mogeneous occupiers. This study does not focus on the differences of occupiers, and the cooperation and games among occupiers are not included as well. Different from single-occupied condition, numerous occupiers exist in a multi-occupied building. In this condition, occupancy is relatively small scale and has a high turnover rate. Therefore, the owner is dominant in the relationship. Occupiers are powerless in negotiation and can only choose to “vote with their feet,” which means relocating to another building. Relocation raises additional costs for occupiers and owner in terms of relocation cost and turnover cost respectively. Table 6 describes these additional costs. Other advantages and disadvantages of green retrofit are similar to those of the condition of single-occupied building. The payoff matrix for owners and occupiers in multi-occupied buildings is shown in Fig. 5, where the superscript “mo” represents the multi￾occupied condition: Based on this adapted matrix, the following scenarios can be investigated to identify the Nash Equilibrium under the multi￾occupied condition. 6.1. The action of the owner when the occupier has the initiative to retrofit In a situation where the owner does not want to invest on retrofit, but the occupiers want to, a single occupier does not have enough economic capabilities or influential power to implement the approach. Therefore, if the active occupiers cannot reach consensus with the reluctant owner and cannot implement retrofit by themselves, they can only choose to “vote with their feet,” which raises the turnover cost for the owner Cmo t21. However, this turnover cost does not increase significantly because small companies consider cost and location as the dominant factors for building selection (Rhoads, 2010). If occupiers are interested in green retrofit, they will support it and pay additional rent to the owner to compensate for the additional costs (Fuerst and McAllister, 2011). Under this condition, the owner can reap a direct payback from the increased rent DRmo ow11 as compensation for the investment. How￾ever, the owner must pay the coordination cost to raise the rent. Coordination cost Cmo c11 is proportional to the number of occupiers. The cost is very high when numerous occupiers, various contracts, and different rental periods exist (Rhoads, 2010), namely, Cmo c11 > > 0. The key information can be summarized as follows: Smo o11 þ DVmo 11 þ DRmo ow11 þ DPmo 11 I mo 11 Cmo c11  0; Cmo t21z0 (9) namely, Bmo ow11  Bmo ow21 (10) According to Formula (10), the owner should choose “reluctant to retrofit” when the initiative of the occupiers is to engage in green retrofit. 6.2. The action of the owner when the occupier is reluctant to retrofit If both owner and occupiers are reluctant to green retrofit, then it is not implemented. The owner obtains neither cost nor profit, such that Bmo ow22 ¼ 0. When the occupier is not active in green retrofit but the owner is, the approach can be implemented without the cooperation of some reluctant occupiers. If some occupiers cannot reach consensus with the owner, they will move out, which results in turnover cost Cmo t12 for the owner. Generally, occupancy rate and rent per unit increase after green retrofit (Ma et al., 2012; Thomas, 2010). Thus, even though some occupiers move out because of retrofit, the owner can reap increased rent DRmo ow12 from the new occupiers. However, high occupancy rate in the future and high rent from a new occupier is an indirect and long-term benefit for the owner. Other profits of retrofit, including operation cost saving Smo o12, increased building value DVmo 12 , and public impact DPmo 12 , are long term and uncertain. By contrast, the costs of retrofit are short term and definite. The risk-adverse investor does not want to take challenges to compare uncertain interests with specific costs. The key information can be summarized as follows: Smo o12 þ DVmo 12 þ DRmo ow12 þ DPmo 12 I mo 12 Cmo t12 Cmo c12  0; Bmo ow22 ¼ 0 (11) namely, Bmo oc12  Bmo oc22 (12) According to Formula (12), the owner should choose “reluctant Table 6 The variables of decision in multi-occupied building. Stakeholder Variable Definition Driver/Barrier Owner Ct Turnover cost of owner BOW1/4 Occupier Cr Relocation cost of occupier BOC6 1308 X. Liang et al. / Journal of Cleaner Production 137 (2016) 1300e1312

X.Liang et aL Journal of Cleaner Production 137(2016)1300-1312 1309 Occupiers Initiative to retrofit Reluctant to retrofit B-SuRAP B2-S0+Aya+AR2+△P Initiative to -Cau --C2-Ccl retrofit Bacu-SeU-D-ARcu Boela=-Cr2 Reluctant to B21=-Cz1 Bm2=0 retrofit B21=-C B2=0 Fig.5.The payoff matrix for owner occupier in multi-occupied building. to retrofit"when the occupier is reluctant to green retrofit. 7.Discussion According to Formulas (10)and (12).the best interests of the owner are served by the "reluctant to retrofit"action regardless of Under the owner-occupied condition,owners can decide by the action occupiers take;thus,"reluctant to retrofit"is the domi- themselves based on Formula (1).However,under the single and nant strategy for the owner(Myerson,2013).Therefore,the action multi-occupied conditions,the Nash Equilibrium is "reluctant to of the occupiers is only discussed in the following section based on retrofit"and "reluctant to retrofit"owing to the interaction be- the situation where the owner is reluctant to retrofit. tween the owners and occupiers at the initial phase.Such re- luctances explain the lack of enthusiasm for green retrofit in the industry.Fig.6 illustrates the mechanism of game theory analysis 6.3.The action of the occupier when the owner is reluctant to under three different occupancy types.The major reasons and is- retrofit sues are discussed below. If both owner and occupier are reluctant to green retrofit,the 7.1.Differences among occupancy types building is operated without green retrofit.Thus,neither cost nor profit exists according to green retrofit,such that B=0.As The difficulty level of green retrofit is strongly related to the mentioned previously,if occupiers want to implement green occupancy type of a building.First,the compensation of investment retrofit but the owner does not,the approach is not implemented, on green retrofit varies according to different occupancy types.In and the active occupiers move out,which incurs relocation feeC. owner-occupied buildings,owners can obtain benefits directly The key information can be summarized as follows: from the green retrofit through low energy consumption.However, S81+AP%-D%-贸≤0，B822=0 in single or multi-occupied buildings,investment of owners in (13) green retrofit will not guarantee direct energy savings.In these conditions,the owners may be rewarded primarily in three other namely, ways.namely,higher rents.lower holding costs and lower risk B21≤B022 (14) (Fuerst and McAllister,2011),compared to owner-occupied build- ings.Retrofit benefits are transferred via the first way from tenants According to Formula(14),the occupier should choose "reluc- to owners.In other words,the green retrofit projects are partly and tant to retrofit"when the owner is reluctant to green retrofit. indirectly funded by the occupiers.The second way,lower holding Based on the above analysis,the Nash Equilibrium for the owner costs,refers to lower maintenance cost and longer operation time and the occupiers is "reluctant to retrofit"and "reluctant to until the next retrofit.The last way involves avoiding premature retrofit,"respectively,under the multi-occupied condition.This obsolescence,policy changing,and risk associated with the future conclusion is consistent with two cases in China,which were increase in energy cost.Among these three rewarding ways,higher studied through interviews conducted in October 2014.One case is rents may be the most direct and promising way to cover costs and the Jin Bin Teng Yue Building,an office building in Guangzhou, earn profits.Risk is difficult to quantify.Low holding cost may not China.The building is occupied by about 400 tenants.The owner be realized due to new technologies and human maintenance be- had the initiative to implement energy-efficient lighting in public haviors during long operation periods.Thus,if owners cannot areas.The negotiation with the tenant committee and tenants with obtain direct energy saving benefits,then they will pursue higher opposite opinions took almost half a year.The facility managers rents or other kinds of direct economic compensation from occu- complained that despite the small-scale of this green retrofit pants.In single-occupied buildings,the single occupier is generally project,coordinating required from them a considerable amount of a large organization that has a long rental contract.If the occupier is time.The other case is the Electronic Technology Building in reluctant to retrofit,raising rent AR in the short term after retrofit Shenzhen,China,which is a commercial building used for the will be difficult.In multi-occupied buildings,the occupiers are in wholesale selling of electronic products.The building has more relatively small scale and have short contracts with owners.Hence than 1000 tenants,many of whom occupy areas less than 10 m. owners can raise rent AR to compensate for the retrofit cost in This building has yet to be green retrofitted.The facility managers the short term. stated they had considered green retrofit,but they had to give up Second,the coordination cost of green retrofit relies on occu- the idea because coordinating more than 1000 tenants was pancy type.Coordinating with occupiers about the green retrofit extremely difficult. decision takes time and incurs labor and economic costs on the part

to retrofit” when the occupier is reluctant to green retrofit. According to Formulas (10) and (12), the best interests of the owner are served by the “reluctant to retrofit” action regardless of the action occupiers take; thus, “reluctant to retrofit” is the domi￾nant strategy for the owner (Myerson, 2013). Therefore, the action of the occupiers is only discussed in the following section based on the situation where the owner is reluctant to retrofit. 6.3. The action of the occupier when the owner is reluctant to retrofit If both owner and occupier are reluctant to green retrofit, the building is operated without green retrofit. Thus, neither cost nor profit exists according to green retrofit, such that Bmo oc22 ¼ 0. As mentioned previously, if occupiers want to implement green retrofit but the owner does not, the approach is not implemented, and the active occupiers move out, which incurs relocation fee Cmo r21. The key information can be summarized as follows: Sso e21 þ DPso 21 Dso 21 I so 21  0; Bmo oc22 ¼ 0 (13) namely, Bso oc21  Bso oc22 (14) According to Formula (14), the occupier should choose “reluc￾tant to retrofit” when the owner is reluctant to green retrofit. Based on the above analysis, the Nash Equilibrium for the owner and the occupiers is “reluctant to retrofit” and “reluctant to retrofit,” respectively, under the multi-occupied condition. This conclusion is consistent with two cases in China, which were studied through interviews conducted in October 2014. One case is the Jin Bin Teng Yue Building, an office building in Guangzhou, China. The building is occupied by about 400 tenants. The owner had the initiative to implement energy-efficient lighting in public areas. The negotiation with the tenant committee and tenants with opposite opinions took almost half a year. The facility managers complained that despite the small-scale of this green retrofit project, coordinating required from them a considerable amount of time. The other case is the Electronic Technology Building in Shenzhen, China, which is a commercial building used for the wholesale selling of electronic products. The building has more than 1000 tenants, many of whom occupy areas less than 10 m2 . This building has yet to be green retrofitted. The facility managers stated they had considered green retrofit, but they had to give up the idea because coordinating more than 1000 tenants was extremely difficult. 7. Discussion Under the owner-occupied condition, owners can decide by themselves based on Formula (1). However, under the single and multi-occupied conditions, the Nash Equilibrium is “reluctant to retrofit” and “reluctant to retrofit” owing to the interaction be￾tween the owners and occupiers at the initial phase. Such re￾luctances explain the lack of enthusiasm for green retrofit in the industry. Fig. 6 illustrates the mechanism of game theory analysis under three different occupancy types. The major reasons and is￾sues are discussed below. 7.1. Differences among occupancy types The difficulty level of green retrofit is strongly related to the occupancy type of a building. First, the compensation of investment on green retrofit varies according to different occupancy types. In owner-occupied buildings, owners can obtain benefits directly from the green retrofit through low energy consumption. However, in single or multi-occupied buildings, investment of owners in green retrofit will not guarantee direct energy savings. In these conditions, the owners may be rewarded primarily in three other ways, namely, higher rents, lower holding costs and lower risk (Fuerst and McAllister, 2011), compared to owner-occupied build￾ings. Retrofit benefits are transferred via the first way from tenants to owners. In other words, the green retrofit projects are partly and indirectly funded by the occupiers. The second way, lower holding costs, refers to lower maintenance cost and longer operation time until the next retrofit. The last way involves avoiding premature obsolescence, policy changing, and risk associated with the future increase in energy cost. Among these three rewarding ways, higher rents may be the most direct and promising way to cover costs and earn profits. Risk is difficult to quantify. Low holding cost may not be realized due to new technologies and human maintenance be￾haviors during long operation periods. Thus, if owners cannot obtain direct energy saving benefits, then they will pursue higher rents or other kinds of direct economic compensation from occu￾pants. In single-occupied buildings, the single occupier is generally a large organization that has a long rental contract. If the occupier is reluctant to retrofit, raising rent DRso ow in the short term after retrofit will be difficult. In multi-occupied buildings, the occupiers are in relatively small scale and have short contracts with owners. Hence owners can raise rent DRmo ow to compensate for the retrofit cost in the short term. Second, the coordination cost of green retrofit relies on occu￾pancy type. Coordinating with occupiers about the green retrofit decision takes time and incurs labor and economic costs on the part Fig. 5. The payoff matrix for owner and occupier in multi-occupied building. X. Liang et al. / Journal of Cleaner Production 137 (2016) 1300e1312 1309