Exploring Key Indicators of Residual Value Risks in China’s Public–Private Partnership Projects Downloaded from ascelibrary.org by Tufts University on 10/28/17. Copyright ASCE. For personal use only; all rights reserved. Jingfeng Yuan1; Wei Xu2; Bo Xia, Ph.D.3; and Mirosław J. Skibniewski, M.ASCE4 Abstracts: Public–private partnerships (PPPs) are gaining popularity in China because of governments’ increasing budget constraints and the urgent need to develop infrastructure since 2014. However, residual value risk (RVR) is a signiﬁcant threat to the success of PPP projects and challenges the governance capability of China’s government, where key risk indicators (KRIs) can be used as a measure to reveal the potential presence, level, or trend of RVR. To help the public sector obtain estimated value as speciﬁed in a concession agreement when the projects are transferred back to the government at the end of the PPP agreement or in earlier termination due to RVR, this paper proposes and reﬁnes a KRI conceptual model composed of seven risk dimensions and 61 indicators. A structured questionnaire survey with PPP experts investigated the relative signiﬁcance of those 61 KRIs for RVR management in PPPs. Although the survey results show that all KRIs are important, seven risk dimensions contribute differently to the exposure of RVR. A conﬁrmatory factor analysis (CFA) was used to test whether the proposed conceptual model ﬁt the observed set of collected data in a predictable way by goodness-of-ﬁt indices, and to further consolidate the KRIs. The reﬁned KRI model uses 41 KRIs based on survey and CFA results, indicating that residual value (RV) of a PPP project is strongly inﬂuenced by effective maintenance, long-term sustainable development, and reasonable proﬁtability and reﬁnanceability. The KRIs including high-quality design work in the early stages of PPP projects, construction quality and public service quality, comparatively low operation costs during the operation period, and feasible technologies adopted by private sectors have a greater value in corresponding dimensions. These 41 KRIs have the ability to indicate trends of losses of RV in PPPs, control weaknesses for RVR management in PPPs, and obtain longterm and sustainable development through PPP projects. DOI: 10.1061/(ASCE)ME.1943-5479.0000561. © 2017 American Society of Civil Engineers. Author keywords: Public–private partnerships (PPPs); Residual value risk (RVR); Key risk indicators (KRIs); Conﬁrmatory factor analysis (CFA); Questionnaire survey. Introduction Public–private partnerships (PPPs) have been adopted widely in China and other countries as signiﬁcant means to provide quality public goods and services through skilled construction and experienced operation by involving the private sector during the concession period (Ng et al. 2010; Chan et al. 2010; Zhang et al. 2015). However, many PPP projects in China are now facing various problems, such as early termination, inadequacy of feasible design, unnormative operation, and fewer investments in or experiences 1 Associate Professor, Dept. of Construction and Real Estate, School of Civil Engineering, Southeast Univ., Nanjing 210096, People’s Republic of China (corresponding author). E-mail: firstname.lastname@example.org 2 Lecturer, Institute of Construction Law, School of Law, Southeast Univ., Nanjing 210096, People’s Republic of China. E-mail: stevenxu@ 126.com 3 Senior Lecturer, School of Urban Development, Queensland Univ. of Technology, Garden Point Campus, 2 George St., Brisbane, QLD 4000, Australia. E-mail: email@example.com 4 Professor, Dept. of Civil and Environmental Engineering, Univ. of Maryland, College Park, MD 20742. E-mail: firstname.lastname@example.org; Ordinary Professor, Institute for Theoretical and Applied Informatics, Polish Academy of Sciences, Gliwice 44100, Poland; T.-S. Yang Honorary Distinguished Professor, Chaoyang Univ. of Technology, Taichung 41349, Taiwan. Note. This manuscript was submitted on February 16, 2017; approved on June 16, 2017; published online on October 27, 2017. Discussion period open until March 27, 2018; separate discussions must be submitted for individual papers. This paper is part of the Journal of Management in Engineering, © ASCE, ISSN 0742-597X. © ASCE with complex infrastructure (Chan et al. 2015; Cheng et al. 2016). As a result, the public sector in China in many cases has limited and ineffective control of PPP projects, which has been evidenced in many prior researches (Zhang 2006; Yuan et al. 2010a; Wibowo and Kochendoerfer 2011; Osei-Kyei and Chan 2015). Of those various problems, residual value risk (RVR) should be viewed as a signiﬁcant threat to the successful implementation of PPP projects in China (Xu et al. 2010). RVR is “the risk that on the expiry or earlier termination of the services contract, the asset (tangible or intangible) is not in accordance with the value (originally estimated by the government), at which the private party agreed to transfer” (Yuan et al. 2015). When RVR occurs, many PPP projects fail to provide quality public goods and services after the concession period elapses or when the PPP project is transferred (Ng et al. 2010; Chan et al. 2010; Li and Zou 2011), which would consequently lead to a high loss of residual value (RV). As a result, the government cannot obtain an estimated value as speciﬁed in the concession agreement when the project is to be transferred back to the government at the end of the PPP agreement or in earlier termination due to RVR. Multiple factors, including project performance, function, proﬁtability, maintainability, operability, sustainability, and the possibility of being reﬁnanced, could inﬂuence the RV (Private Finance Panel 1996; Burke and Demirag 2015; Yuan et al. 2015). RVR could result in high maintenance cost, low-quality service, and functional problems of facilities, which would fail to meet the prescribed requirements in PPP contracts. In this case, the accountability and credit worthiness of public sectors must be greatly challenged (Rwelamila et al. 2015; Chan et al. 2015). Moreover, RVR may harm the beneﬁts of the general public (Yuan et al. 2015; 04017046-1 J. Manage. Eng., 2018, 34(1): 04017046 J. Manage. Eng. Downloaded from ascelibrary.org by Tufts University on 10/28/17. Copyright ASCE. For personal use only; all rights reserved. Bulsara et al. 2015). In addition, the RVR will damage the collaboration between public and private sector, forcing the private sectors to quit future PPP market opportunities. Therefore, effective RVR management is critical for the success of PPP projects, comprising the identiﬁcation of critical factors leading to RVR, evaluation of RVR, and precontrol of RVR. In particular, an accurate identiﬁcation of the critical factors leading to RVR and key risk indicators (KRIs) of RVR would greatly improve RVR management. Furthermore, the reason for RVR has been mainly concluded as the cumulative effects on the change of RV in PPP projects (Algarni et al. 2007; Yuan et al. 2016). Six critical risk factors have been identiﬁed by authors’ prior work: (1) downfall of product or service performance, (2) functional problems, (3) decrease of proﬁtability and low possibility of being reﬁnanced, (4) deterioration of maintainability, (5) decline of operability, and (6) failure of sustainability (Yuan et al. 2015). However, KRIs for those critical risk factors that measure how risky an activity is have not yet been thoroughly identiﬁed. KRIs can be utilized as tools for monitoring controls, risk drivers, and exposures. KRIs can shed light on the potential risk events. Therefore, this study aimed to develop an effective set of KRIs for China’s PPP projects. The ﬁndings will provide useful insights about potential risks that may have an undesirable impact on the achievement of appropriate and effective RVR management, which can help China’s government improve its ability to reduce RVR and provide better public facilities and services. The remainder of this paper begins with a literature review of the related ﬁelds, followed by a presentation of the research approaches used in the study, and establishment of the proposed conceptual model of KRIs for RVR management for PPP projects to describe theoretical relationships between KRIs and six critical factors. After that, a national research survey aimed at investigating KRIs in China’s PPP projects is demonstrated. A series of statistical methods were adopted to analyze the data, and a conﬁrmatory factor analysis (CFA) was used to test whether the conceptual model ﬁt the observed set of collected data as predicted through ﬁt indices and improve the model, identifying KRIs and clarifying their relationships. RVR Management and Related KRIs in PPPs Literature Review Recent Development of PPPs in China In the past three decades, the demands for infrastructure projects in China have been largely driven by urbanization (Zhang et al. 2015). Owing to the limited ﬁnancial budget and inefﬁcient experiences in providing infrastructure products and services involving solely the public sector, Chinese policymakers must ﬁnd new ways to improve the performance of public projects and services to meet the public’s demands (Osei-Kyei and Chan 2015). Because of PPPs’ attractive characteristics of transferring risks to private partners, reducing public sector administration costs, solving the problem of public sector budget restraint, providing higher-quality public products and services, and saving time in delivering the projects, PPPs have been gaining popularity in recent years in China to involve greater private sector participation in infrastructure (Zhang et al. 2015). Meanwhile, a variety of PPP studies have been conducted in the ﬁelds of procurement (Lam and Chow 1999; Liu et al. 2017), risk management (Akintoye and Chinyio 2005; Cheung and Chan 2011), critical success factors (Chou et al. 2012; Zhao et al. 2013; Osei-Kyei and Chan 2015), driving factors (Chan et al. 2009; Yuan et al. 2010a), decision making (Jin 2010; Yuan et al. 2010b; Wibowo et al. 2012; Xu et al. 2012; Juan et al. 2016), and © ASCE stakeholder management (Appuhami et al. 2011; Brinkerhoff and Brinkerhoff 2011). In PPPs, both public and private sectors should make efforts in ﬁnancing, project planning and design, construction management, maintenance, facility management, staff training, and technology progress to achieve value for money (Grimsey and Lewis 2005; Sobhiyah et al. 2009). PPP projects with a typical concession period from 15 to 30 years are usually under the operation of the private sector. In China, new issues and lessons emerge every day. Many prior works have indicated that government decision making, government creditworthiness, law and policy, inﬂation, and price change would greatly impact the success of PPP projects in China (Song et al. 2013; Zhao et al. 2013; Chan et al. 2015). Hence, the private sectors in China’s PPP projects ultimately have to face huge challenges. Meanwhile, the governance of projects has been reported as weak and ineffective in many cases due to limited control and monitoring by China’s public sector (Yuan et al. 2010a; Wibowo and Kochendoerfer 2011; Zhang 2006). Therefore, RVR could arise when the project is to be transferred back to the government at the end of the PPP agreement or in earlier termination because the estimated value cannot be achieved (Yuan et al. 2015). Since 2014, the central government of China has facilitated the process of PPP projects, attempting to make a transition from user-pay PPP to governmentpay PPP systems with the aim of facilitating infrastructure investment and government transformation (State Council of China 2014; Chang and Chen 2016). In this case, the governments in China would be more concerned about the effectiveness and efﬁciency of PPP project transfer (Liu et al. 2016), which means the RVRs in China are becoming more and more signiﬁcant. Information from prior researches or case studies demonstrates that many disputes and losses related to residual values and the transferring of PPP projects can be attributed to policy and law change, design problems, poor maintenance of the assets, completion delay, high costs and prices, and public opposition (Infrastructure Australia 2008a; Xu et al. 2010; Burke and Demirag 2017). The reason can be concluded that the public sector cannot obtain the return in the desired condition (Chan et al. 2011; Yuan et al. 2015, 2016). RVR has been identiﬁed as a critical risk that will have a strong inﬂuence on the value of initial bids and the incentives facing contractors (Private Finance Panel 1996; Hall 1998; Froud 2003; Algarni et al. 2007; Jin and Zhang 2011). However, many public sectors may neglect the impacts of RVR because most PPP projects have a long concession period before they are transferred to public sectors. Hall (1998) demonstrates that risks associated with how much the assets will be worth when a PPP project is transferred to public sectors are very important. Many problems, such as high maintenance cost, low-quality service, and functional problems of the facilities, will occur if RVR cannot be identiﬁed, evaluated, or managed. Demirag et al. (2011) believed that RVRs were the most important in the context of PPP because they were most likely to affect the goal of off-balance sheet project treatment. Therefore, suitable management of RVR would be of tremendous value for better maintenance, easier transfer of PPP projects, improvement of ongoing management, and sustainable development of the infrastructure (Hall 1998). Yuan et al. (2015) indicated that RVR in PPPs can be described as the uncertainty that the RV could be lower than the estimated or anticipated value determined in the PPP contract when being transferred. From the perspective of uncertainty, a traditional risk-management framework can provide useful 04017046-2 J. Manage. Eng., 2018, 34(1): 04017046 J. Manage. Eng. Downloaded from ascelibrary.org by Tufts University on 10/28/17. Copyright ASCE. For personal use only; all rights reserved. lessons for RVR management. Usually, a risk-management framework includes risk identiﬁcation, risk evaluation, risk allocation, and risk control or treatment. Many previous studies have focused on the RVR management in PPPs. Algarni et al. (2007) indicated that RVR could occur when facility assets have been suffering from years of neglect, overuse, deferred maintenance, and delayed repair, which would harm the sustainable development of the infrastructure; this is agreed by Xu et al. (2010) and Chan et al. (2011). Many ofﬁcial documents have identiﬁed related risk factors of RVR, such as construction costs, total service demand, total operating costs, and technological change (Partnerships Victoria 2001; Infrastructure Australia 2008a, b; Comptroller and Auditor General of India 2009). How to allocate RVR between public and private sectors has also been discussed (Arndt 1998; Li et al. 2005; Ke et al. 2010; Marques and Berg 2011; Krüger 2012). Froud (2003) proposed a RVR treatment method, which deliberated whether RVR can be dealt with by simply writing off the asset completely over the life of the contract. Risk identiﬁcation is the ﬁrst step for risk management (PMI 2013). The principal work of risk identiﬁcation is to detect the factors that can potentially jeopardize the successful conclusion of a PPP project by causing cost overruns, time delay, and low level of performance compared to the output speciﬁcation (Carbonara et al. 2015; Liu et al. 2016). Many prior studies have identiﬁed risk factors in PPP projects in China, India, the United States, the United Kingdom, etc., and risk factors of different PPP project types, such as transportation, water treatment, power plant, and education (Lam and Chow 1999; Wang et al. 1999; Thomas et al. 2006; Chan et al. 2011; Zhang et al. 2015; Ameyaw and Chan 2015a). Generally, government-related risks and assets-related risks are the most important risks for PPP projects. These studies strengthened the perceptions of PPP risks and explored appropriate ways to fully understand signiﬁcant risks in PPPs. However, KRIs, which can be used as measures to reveal the potential presence, level, or trend of a risk, have not been developed for RVR management. KRIs can indicate whether a risk has occurred or is emerging, the level of risk exposure, and the trend of changes in risk exposure. Therefore, KRIs can provide information about a risk situation that may or may not exist and serve as a signal for further action (Beasley et al. 2010). When implemented properly, KRIs can provide signiﬁcant insight into changes in the risk proﬁle and bring additional strategic and operational value (Fraser and Simkins 2011; Zhang 2011; Huang et al. 2012). KRI has been adopted widely in the ﬁelds of ﬁnancing, enterprise management, banking, and e-business (Heckmann et al. 2015; Juan et al. 2016). In the context of RVR, KRIs can be used as a tool to predict the change of RVR and allow for proactive intervention before the project is transferred (Yuan et al. 2015, 2016). Therefore, a signiﬁcant work of this paper is to develop KRIs for RVR in China’s PPP projects by analyzing the events that affected the RV of PPP projects in the past or the present and go back to the roots of RV to identify and fully understand the underlying cause of the RV. Knowledge Gap According to the review of previous studies, the following knowledge gaps can be identiﬁed. First, the lack of effective RVR management leads to many problems during the process to reduce the RV of PPP projects. Considering that long-term vision on PPP projects is essential for pursuing public beneﬁts, the RV of PPP projects should be kept at a high level, and RVR management should be constantly improved to realize public beneﬁts. Second, precisely knowing which indicators are to be used to measure and indicate risk factors that lead to RVR is another knowledge gap. The usage of KRIs can contribute to improving not only RV of PPP projects, but also project performance through a proactive management position instead of a reactive one (Beasley et al. 2010). At the same time, the interrelationships between KRIs are crucial to RVR management and should be further clariﬁed. This paper attempts to ﬁll these knowledge gaps, with particular emphasis on the second knowledge gap. Identiﬁcation of KRIs would be very useful for establishment of RVR management framework, partially addressing the ﬁrst knowledge gap. Research Method In this study, a series of qualitative and quantitative research methods were used to explore the conceptual KRI system of RVR to measure and manage RVR in PPPs effectively, as shown in Fig. 1. First of all, a conceptual model of KRIs in PPPs was proposed according to KRIs’ theoretical relationships. Based on the identiﬁed factors that led to RVR from authors’ prior work, the authors developed an indicator system to measure RVR and identiﬁed KRIs in PPP projects. A survey of 138 PPP experts in China using a stratiﬁed random sampling method with 5-point Likert-style rating questions was conducted to investigate the KRIs of a PPP project during its lifecycle in China. Statistical analyses were performed, using the SPSS 17.0 software package, by multiple methods, including Cronbach’s alphas, mean value, and grouping discussion to validate the survey, as well as to evaluate the perceptions of risk indicators. A CFA was performed using the LISREL 8.70 software to test whether the theoretical model ﬁt the survey data, and to clarify the relationships among KRIs. In addition, the insigniﬁcant indicators were removed from the initial indicator system based on signiﬁcance level and factor loadings. Then, the KRIs were identiﬁed for RVR in PPPs. The CFA aims not only at detecting and explaining but also improving the present state of RVR management in real PPP projects and developing helpful risk management in PPPs for future use. Fig. 1. Methodology adopted in this study © ASCE 04017046-3 J. Manage. Eng., 2018, 34(1): 04017046 J. Manage. Eng. Conceptual Model of KRI System for Measuring RVR in PPPs Downloaded from ascelibrary.org by Tufts University on 10/28/17. Copyright ASCE. For personal use only; all rights reserved. Proposed Conceptual Model In prior studies, six dimensions of internal risk factors have been identiﬁed according to their impacts on the RV change of PPP projects (Yuan et al. 2015). Internal risk factors, called local risk factors (LRFs), come from the internal environment, and are relatively more controllable within PPP projects and will vary between projects. In addition, RV would inevitably be inﬂuenced by the relatively uncontrollable external environment and systematic impacts (Frilet 1997; Wang et al. 2000), which can be called global risk factors (GRFs). All GRFs would affect PPP projects by combining with the inﬂuence of LRFs. Therefore, the RVR factors can be divided into LRFs and GRFs. According to related researches (Beasley et al. 2010; Fraser and Simkins 2011), each risk factor can be measured through a series of KRIs. Thus, KRIs can be identiﬁed with the help of a proposed conceptual model comprising all risk factors of RVR in PPPs, as shown in Fig. 2. KRIs may represent key ratios that management tracks as indicators of evolving RVR, which points out the need for actions to be taken. Local key risk indicators (LKRIs) are identiﬁed from six aspects: (1) downfall of product or service performance, (2) functional problems, (3) deterioration of maintainability, (4) decline of operability, (5) failure of sustainability, and (6) decrease of proﬁtability and low possibility of being reﬁnanced. In addition, global key risk indicators (GKRIs) are identiﬁed from the external environment and systematic impacts on RV and other internal KRIs. A conceptual model of the KRI system for measuring RVR in PPPs is shown in Fig. 2. Identification of Potential KRIs During the process of PPP projects, many internal and external factors could inﬂuence the change of RV in PPPs during the concession period; these factors are LRFs and GRFs. • PPP projects usually suffer great and continuous changes from the stage of preimplementation to the stage of posttransfer due to internal factors within PPP projects. Hence, the RV of PPPs would be inﬂuenced in different stages by many different internal risk factors, which are LRFs. Therefore, the LKRIs for RVR in PPPs should be selected to consider and assess the full range of LRFs that could potentially harm RV or lead to missed performance objectives in any stages of the concession period. To identify LKRIs for RVR in PPPs, the interaction of process viewpoint and residual value change should be carefully considered. Usually, the process of a PPP project before being transferred to public sectors can be divided into ﬁnancing, design, construction, and operation. Meanwhile, the six dimensions of risk factors (D1–D6) could impact on the RV in different stages. Thus, the LKRIs for RVR in different stages in PPPs can be further identiﬁed according to the effects of different internal risk dimensions, as shown in Table 1. • The GRFs contain the changes in political, social, legal, and economic conditions of a speciﬁc PPP project as well as any changes in contract, force majeure, and interface. GRFs are exacerbated by interdependencies among the different units often because of strong links between PPP systems and external environments or between different units within the PPP system. These risks can be triggered by sudden events or events built up over time, with the impact often being large and possibly catastrophic. The GKRIs can be used to measure the inﬂuence of GRFs with potential losses or damages to an entire system as contrasted with losses to a single unit of that system. GKRIs for RVR in PPPs can be identiﬁed according to prior researches on the risk management of PPP projects concerning the impacts from macro and external environment as well as government-related inﬂuences, which are different perspectives within the dimension of GRFs (D7) as shown in Table 2. The RVR could be fully understood and thoroughly analyzed when GKRIs and LKRIs are integrated and then used together. Thus, the potential KRI lists can be obtained according to the proposed model in Fig. 2, a literature review on PPPs and KRIs, and the process viewpoint, which are shown in Table 1. The detailed description of different dimensions of KRIs (D1–D7) are shown as follows. D1: Indicators Related to Performance In this dimension, the identiﬁed LKRIs should be used to measure the downfall of product or service performance delivered by PPPs. From the perspective of public goods or services provided by PPP projects, the requirements for public facilities mean high quality as well as timely and convenient service (Li et al. 2005; Yuan et al. 2009), which strongly inﬂuences the tangible assets of RV. The decline of performance would greatly inﬂuence the RV of PPPs, Fig. 2. Conceptual model for KRIs for residual value of PPP projects © ASCE 04017046-4 J. Manage. Eng., 2018, 34(1): 04017046 J. Manage. Eng. Table 1. Identiﬁcation of Potential LKRIs Based on Process-Based Viewpoint Local risk factors (D1–D6) Stage Financing Design Downloaded from ascelibrary.org by Tufts University on 10/28/17. Copyright ASCE. For personal use only; all rights reserved. Construction Operation Process-based possible influences on the change of RV Reasonability of project ﬁnancing structure Project self-liquidation Availability of ﬁnancing Financing costs Design reasonability or major changes Delay of construction Overrun of construction costs Construction quality Supply of goods and materials Level of advanced technology Public service quality Operation costs Financial capability Market stability Reasonability of pricing Maintenance safety Reasonability of ownership Maintenance costs Environmental protection Reasonability of organizational structure Horizontal competition D1 Performance D2 Function D3 Maintainability D4 Operability D5 Sustainability D6 Profitability and refinanceability 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 = potential LKRIs are from the impacts from Dx (x = 1,2, …, 6) on the change of RV in different stages. Table 2. Identiﬁcation of Potential GKRIs Based on External and Systematic Impacts Global risk factor (D7) Political impacts Social impacts Legal impacts Economic impacts Systematic impacts Potential risk indicator Effectiveness of policies (e.g., ineffective supervision and corruption) Change of government ofﬁcials Creditworthiness of different stakeholders Public opposition Legal risks (e.g., improper laws and rules, regular changes of laws and rules) Change of interest Change of exchange rates Inﬂation Imperfect ﬁnancing market Decision-making capability Land acquisition Project approval Contract management Interface management Force majeure which could be indicated by the design reasonability and quality, construction quality, construction duration, construction cost, technology progressiveness, operation safety, and maintenance cost. D2: Indicators Related to Function In this dimension, the identiﬁed LKRIs should be used to measure whether a series of facilities, equipment, instruments, and technical © ASCE or management documents used in a PPP project can meet the function requirements and can ensure that high performance will be achieved. Ravindran (2010) demonstrated that necessary function traditionally was the basis for providing quality goods and services for a public facility. The indicators of functional problems can be proposed as design reasonability and quality, construction quality, construction duration, public service quality, organization change, maintenance cost, operation cost, price adjustment, and supply efﬁciency and effectiveness. D3: Indicators Related to Maintainability In this dimension, the identiﬁed LKRIs should be used to measure whether the PPP project can be carefully maintained to isolate defects or their causes, correct defects or their causes, meet new requirements, make future maintenance easier, or cope with a changed environment (Li et al. 2005; Chan et al. 2009). Furthermore, the reliability and serviceability of infrastructure projects are both of importance for RVR management in a PPP project (Yuan et al. 2015). Therefore, the indicators to measure the deterioration of maintainability should include reasonable planning and design, effective quality and cost control, and technology improvement. D4: Indicators Related to Operability In this dimension, the identiﬁed LKRIs should be used to measure the ability to keep a system or a whole PPP project in a safe and reliable functioning condition, according to predeﬁned operational requirements. Meanwhile, LKRIs should also reﬂect the decline of operability of RV in PPPs due to changes of tangible and intangible assets, which would include satisfaction of stakeholders, change of subsystems, interrelationships among different project phases or 04017046-5 J. Manage. Eng., 2018, 34(1): 04017046 J. Manage. Eng. Downloaded from ascelibrary.org by Tufts University on 10/28/17. Copyright ASCE. For personal use only; all rights reserved. contracts, advanced technologies, and efﬁcient project organization (Sharma 2007). The changes could be derived from low public service quality, operation cost overrun, market inﬂation, horizontal competition, pricing adjustment, maintenance accidents, technology improvement, and unclear ownership. D5: Indicators Related to Sustainability In this dimension, the identiﬁed LKRIs should be used to measure social, environmental, and ﬁnancial impacts on the change of RV of PPPs (Koppenjan and Enserink 2009). The LKRIs for sustainable RV management of PPPs should indicate whether the process of social development, utilizing resources, directing investment, and institutional change can enhance both current and future potential to meet present and future needs after a PPP project is transferred to public sectors (Mirza 2006). Social, environmental, and ﬁnancial sustainability in PPPs can provide a good basis for long-term realization, maintenance, and operation of public infrastructures (De Los Ríos-Carmenado 2016). The uncertainty of sustainability could be indicated by design reasonability and quality, construction quality, technology progressiveness, maintenance cost, public service quality, and environment protection. D6: Indicators Related to Profitability and Refinanceability In this dimension, the identiﬁed LKRIs should be used to measure ﬁnancial conditions. The LKRIs for proﬁtability and reﬁnanceability for RVR in PPPs can indicate the potential of tangible and intangible assets in PPP projects to be invested after being transferred to public sectors. Related indicators include ﬁnancing structure, the project’s self-liquidating ratio, availability of ﬁnancing, and capital costs, which are important for proﬁtability of PPP projects and have a strong relationship with ﬁnancing issues. Furthermore, the impacts on proﬁtability will inﬂuence the process of construction and operation, which would lead to a lower possibility of being reﬁnanced when a PPP project is transferred to public sectors (Xenidis and Angelides 2005; Wibowo et al. 2012). D7: Indicators Related to Global Risks Obviously, a PPP project is usually widely inﬂuenced by many macro and external risk factors. According to prior researches (Wang et al. 2000; Akintoye and Chinyio 2005; Ameyaw and Chan 2015b; Carbonara et al. 2015), the inﬂuences have a strong relationship with political, social, legal, economic, systematic, and other impacts. Political risks in PPPs could result from change of sovereignty and policies, unstable regime system, weak supervision system, and corruption. As presented by Wang et al. (1999) and Yuan et al. (2010a), the commitments between public and private sectors are critical to the success of PPP projects, where the degree of trust between public and private sectors is signiﬁcant. For example, the possibility of local authorities living up to the ﬁnancial obligations can attract more investments in PPP projects in the long run (Witz et al. 2015). Meanwhile, support from the general public would reduce the social risks in PPPs (Rouhani et al. 2016). Economic inﬂuences would inevitably play key roles in PPP projects and make the RV change, where the inﬂations of interest, exchange rate, and currency are closely related (Wang et al. 2000). Moreover, the imperfect ﬁnancing market in China also has negative impacts on the RV change of PPP projects (Wang et al. 2000; Chan et al. 2011; Yuan et al. 2015). In fact, many systematic risks can be included into this dimension because they can affect the PPP projects during their whole lifecycle. These indicators can be classiﬁed into two types. One is the work in the early stage of PPP projects that is very important, such as decision making, land acquisition, and project approval (Xu et al. 2012; Juan et al. 2016). Another is the work that should be well implemented during the process, such as contract © ASCE management and interface management for different stages and different stakeholders in PPPs (Klijn and Koppenjan 2016; Liu et al. 2017). Furthermore, force majeure is usually ignored in PPP projects, which could greatly harm or reduce the RV of PPPs (e.g., earthquake, ﬂood, volcanic eruption) (Carbonara et al. 2015; Xu et al. 2015). Hypothesized Relationships in the Conceptual Model A conceptual framework (Fig. 2) is a representation for the RVR in PPP projects, and serves as the foundation for exploring KRIs. PPPs are focused on delivering the value of public products and services, including the satisfaction of the general public, efﬁciency and effectiveness, future potential of projects, and performance improvement. Therefore, any changes of PPP projects, which inﬂuence the value of public products and services, would inevitably reduce the RV of PPP projects. Based on these relationships, the following hypothesis can be concluded. The conceptual model can be used to measure the RVR of PPP projects, where all risk dimensions can contribute to the change of RV in PPP projects. As shown in Fig. 2, all risk dimensions contribute to the RV in PPP projects from different perspectives. Although their contributions and pathways are different, they all inﬂuence the RV of projects. Furthermore, the classiﬁcation of these dimensions within the proposed model speciﬁcally reﬂects the characteristics of RV in PPPs. Additionally, the survey data should provide an empirical evaluation of the arrangement of indicators into the different dimensions. Research Survey on KRIs for RVR Introduction of Survey To obtain the opinions of professionals and researchers on KRIs for RVR and their differences, a structured questionnaire survey was conducted at the annual conference of The Architectural Society of China–Construction Management Research Sector (ASC-CMRS) in August 2015. The survey targeted individuals and organizations with experience in PPPs. The questionnaire comprises three parts. The ﬁrst part seeks background information about the respondents and general issues about PPPs (Part A). The second part deals with KRIs for RVR in PPP projects (Part B). In this part, the investigation of different KRIs for RVR within PPPs was conducted by using Likert-style rating questions with a 5-point scale to identify the relative significance of KRIs for RVR, by which the respondents’ opinions about the importance of each indicator can be elicited. In the questionnaire, an initial conceptual model of KRIs for RVR in PPPs based on a literature review was presented. The respondents were asked to provide their full opinions on the importance of KRIs base on the conceptual model, through which the level of agreement or disagreement generally can be measured. The scale intervals are interpreted as follows: (1) can be ignored or not important, (2) maybe important, (3) important, (4) very important, and (5) most important. A total of 315 questionnaires were sent out at the conference site, of which 138 were completed and returned. The effective return rate was 44% compared to acceptable rates in prior studies of PPPs, including 36% in the Chan et al. (2011) study and 11.4% in the Jin and Zhang (2011) study. The response was therefore deemed adequate for the purposes of data analysis. The respondents were from construction enterprises, research organizations, and consulting companies. All participants met the following requirements: (1) extensive work experience within the construction industry, and (2) 04017046-6 J. Manage. Eng., 2018, 34(1): 04017046 J. Manage. Eng. involved in the management of PPP projects or gained in-depth knowledge of the PPP model through research. The survey data show that 52.7% of respondents were from academia, and 47.3% of respondents were from industry. The experiences of respondents can be found in Tables 3 and 4. Table 3 describes the experiences of respondents in the construction industry and PPPs. Table 4 shows the PPP project types reﬂected in respondents’ experience. Downloaded from ascelibrary.org by Tufts University on 10/28/17. Copyright ASCE. For personal use only; all rights reserved. Consistency of Survey Data A reliability analysis was conducted to test the internal consistency of the survey variable data. Cronbach’s alphas were 0.846 for all LKRIs and 0.922 for all GKRIs. The values are much higher than 0.70 as recommended in the guideline by Nunnally (1978), which suggests that in the early stages of research on predictive tests or hypothesized measures of a construct, reliability of 0.70 or higher should sufﬁce. Cronbach’s value was derived for each indicator. Values exceeding 0.9, between 0.9 and 0.7, between 0.7 and 0.6, between 0.6 and 0.5, and lower than 0.5 indicate excellent, good, acceptable, poor, and unacceptable reliability, respectively. The opinions of different participants on all KRIs were similar based on the value of Cronbach’s alphas, which provides a basis for development of further studies. The values of Cronbach’s alphas as mentioned earlier are the average values of Cronbach’s alpha over all LKRIs and GKRIs, which are higher than 0.9 or close to 0.9. Therefore, it is not necessary to display all alpha values derived for each KRI (Bonett and Wright 2015). Actually, many prior studies also only used the average values of Cronbach’s alpha to indicate the consistency of survey data (Chou et al. 2012; Tang and Shen 2013; Seshadhri and Topkar 2016). Description of Survey Results Related to LKRIs The survey results of LKRIs are shown in Table 5. The mean response rating values for the 46 LKRIs range from a maximum of Table 3. Survey Respondents’ Related Experiences in Industry Experience ≤5 years 6–10 years 11–15 years 16–20 years ≥21 years Total In construction industry Percentage In PPPs Percentage 13 15 56 49 5 138 9.42 10.87 40.58 35.51 3.62 100.00 53 47 19 12 2 22 38.41 34.06 13.77 8.69 9.09 100.00 Table 4. PPP Project Types Reﬂected in Survey Respondents’ Experience Project type Hospital Transportation Water and sanitation Power and energy Housing Sports Environmental protection IT and communication School and education Urban development Others Total © ASCE Number Percentage 12 74 49 16 41 9 26 21 23 31 5 307 3.91 24.10 15.96 5.21 13.36 2.93 8.47 6.84 7.49 10.09 1.63 100 4.545 (D3.1, design reasonability and change in the maintenance dimension) to a minimum of 2.727 (D2.3, construction duration in the function dimension). No indicator mean value scores fell into the not important (<1.5) category, which indicates that 45 LKRIs are considered important and can be used to measure the RVR in PPPs. The mean values of all LKRIs from six dimensions also indicate that the identiﬁed KRIs can be used to analyze the RVR from six perspectives. Meanwhile, some indicators can measure different risk factors at the same time, such as design reasonability and change, which simultaneously serves as the dimension of performance, function, maintainability, and sustainability, and can be described as the impacts of different risk dimensions. These indicators actually contribute differently in different dimensions, which moreover illustrates that the occurrence of RVR in PPPs is led by interactions between different risk factors (dimensions). In the survey, factors from each of the dimensions were investigated. Thus, a group discussion of the different dimensions is necessary. The rankings of risk factors in the corresponding dimensions are shown in Table 5. Detailed analysis within different dimensions can be provided as follows. Dimension of Performance Based on the survey results, the KRIs in this dimension should include indicators that reﬂect the impact of designing work (D1.2), quality and technologies (D1.2, D1.5), traditional performance indicators (D1.3, D1.4), and maintenance (D1.6, D1.7). Because performance for RV management places great emphasis on goods and services delivered by PPPs (Clifton and Dufﬁeld 2006), the indicators related to quality are more signiﬁcant than other traditional performance indicators (e.g., schedule and costs). The reasonability of design work is the most important in this dimension (D1.1, 4.500), which would also inﬂuence the quality and technologies to be selected by the concessionaire. In truth, designing work involves planning for quality, which can provide excellent support for construction and operation (Yuan et al. 2010b). In recent years, performance-based design, which is an engineering approach to design elements of infrastructure based on agreed upon performance goals and objectives, engineering analysis, and quantitative assessment of alternatives against the design goals and objectives using accepted engineering tools, methodologies, and performance criteria, has become more and more important in the ﬁeld of PPP. As presented by Love et al. (2015), performance-based design can lead to innovation and new methods. Thus, more performance-based design for infrastructure in PPPs would notably improve the RV of PPP projects. Dimension of Function Ravindran (2010) indicated that necessary function traditionally was the basis for providing quality goods and services for a public facility. To meet the function requirements to ensure that high performance can be achieved, the indicators related to design (D2.1, 4.409) and quality (D2.2, 4.091) are more signiﬁcant compared with other indicators based on the survey data, because the improvement of design quality and construction quality would reduce functional problems (Dikmen et al. 2005). In addition, the operation stage is important for improving function of facilities, which is reﬂected by the score of public service quality (D2.4, 3.455), maintenance costs (D2.6, 3.227), operation costs (D2.7, 3.591), and project pricing (D2.8, 3.409). Obviously, notwithstanding the early stage of a PPP project can determine the function of the whole project, heavy work actually should be done during the operation period to ensure the basic and important functions are all kept at a high level to deliver quality goods and services. 04017046-7 J. Manage. Eng., 2018, 34(1): 04017046 J. Manage. Eng. Table 5. LKRIs for RVR in PPP Projects and Their Scores and Rankings in Different Dimensions Downloaded from ascelibrary.org by Tufts University on 10/28/17. Copyright ASCE. For personal use only; all rights reserved. Distribution shape Local key risk indicator Mean D1 Performance D1.1 Impacts of design on performance D1.2 Impacts of construction quality on performance D1.3 Impacts of construction delay on performance D1.4 Impacts of construction costs overrun on performance D1.5 Impacts of level of advanced technology on performance D1.6 Impacts of maintenance safety on performance D1.7 Impacts of maintenance costs overrun on performance D2 Function D2.1 Impacts of design on function D2.2 Impacts of construction quality on function D2.3 Impacts of construction delay on function D2.4 Impacts of service quality on function D2.5 Impacts of organizational structure on function D2.6 Impacts of maintenance costs overrun on function D2.7 Impacts of operation costs overrun on function D2.8 Impacts of pricing overrun on function D2.9 Impacts of goods and materials supply on function D3 Maintainability D3.1 Impacts of design on maintainability D3.2 Impacts of construction quality on maintainability D3.3 Impacts of level of advanced technology on maintainability D3.4 Impacts of maintenance costs overrun on maintainability D4 Operability D4.1 Impacts of service quality on operability D4.2 Impacts of operation costs overrun on operability D4.3 Impacts of horizontal competition on operability D4.4 Impacts of market stability on operability D4.5 Impacts of pricing on operability D4.6 Impacts of maintenance safety on operability D4.7 Impacts of level of advanced technology on operability D4.8 Impacts of ownership on operability D5 Sustainability D5.1 Impacts of design on sustainability D5.2 Impacts of construction quality on sustainability D5.3 Impacts of level of advanced technology on sustainability D5.4 Impacts of maintenance costs overrun on sustainability D5.5 Impacts of service quality on sustainability D5.6 Impacts of environmental protection on sustainability D6 Proﬁtability and reﬁnanceability D6.1 Impacts of ﬁnancing structure on proﬁtability and reﬁnanceability D6.2 Impacts of project self-liquidation on proﬁtability and reﬁnanceability D6.3 Impacts of availability of ﬁnancing on proﬁtability and reﬁnanceability D6.4 Impacts of ﬁnancing costs on proﬁtability and reﬁnanceability D6.5 Impacts of construction delay on proﬁtability and reﬁnanceability D6.6 Impacts of construction costs overrun on proﬁtability and reﬁnanceability D6.7 Impacts of operation costs overrun on proﬁtability and reﬁnanceability D6.8 Impacts of horizontal competition on proﬁtability and reﬁnanceability D6.9 Impacts of market stability on proﬁtability and reﬁnanceability D6.10 Impacts of pricing on proﬁtability and reﬁnanceability D6.11 Impacts of maintenance costs overrun on proﬁtability and reﬁnanceability D6.12 Impacts of level of advanced technology on proﬁtability and reﬁnanceability 3.649 4.500 4.318 3.000 3.045 3.909 3.364 3.409 3.511 4.409 4.091 2.727 3.455 3.182 3.227 3.591 3.409 3.500 4.136 4.545 4.273 3.909 3.818 3.665 3.955 4.091 3.591 3.591 3.636 3.500 3.409 3.545 3.902 4.409 4.136 3.773 3.682 3.636 3.773 3.826 3.955 4.182 4.045 3.864 3.318 3.773 4.091 3.909 3.955 3.773 3.727 3.318 Dimension of Maintainability Effective maintainability aims at substantially improving reliability and serviceability for facilities in PPP projects, eventually reducing RVR. As presented before, performance-based design focuses not only on performance improvement, but also on hindering deterioration of maintainability in advance through careful and innovative © ASCE SD 0.5976 0.7162 1.1547 0.9989 0.8679 0.9535 0.8541 0.666 0.868 0.935 1.101 0.795 0.813 0.796 1.054 0.802 0.596 0.631 0.811 0.907 0.722 0.868 0.854 0.854 0.79 0.74 0.734 1.143 0.734 0.889 0.813 1.086 0.953 1.066 1.09 0.795 0.785 0.889 0.945 0.922 0.75 0.971 1.09 0.685 0.767 0.839 Rank 5 1 2 7 6 3 5 4 6 1 2 9 4 8 7 3 5 6 1 1 2 3 4 4 2 1 4 5 3 7 8 6 2 1 2 3 4 5 3 3 4 1 3 7 12 9 2 6 5 8 10 11 Skewness Kurtosis –0.736 –0.569 0.000 –0.413 0.187 0.249 0.058 –0.312 –0.756 –0.771 0.223 –1.687 –0.683 –0.399 –0.699 –0.667 0.222 –0.109 0.274 0.126 0.296 –0.140 0.000 –0.429 –0.102 0.786 –1.300 –0.080 –0.358 –0.380 –1.215 –0.202 –0.933 –0.269 0.175 –0.453 0.025 –0.463 –1.437 –0.308 0.069 –0.667 –0.058 –0.058 –0.496 0.388 0.742 –0.438 –0.929 –0.102 –0.399 –0.399 0.182 –0.019 0.415 –0.428 –0.847 –0.734 –0.126 –0.762 0.114 –0.280 –0.538 –0.179 –0.358 0.346 –0.934 –1.141 –0.631 –0.352 –0.083 –0.161 –0.349 –0.305 –0.154 –0.490 –0.631 0.323 –0.167 0.372 –0.884 –1.292 –1.319 –0.844 –1.217 –0.567 –1.106 –0.641 –0.884 –0.697 –0.044 –0.090 design (Love et al. 2015). Thus, design-related indicators obtain the highest score (D3.1, 4.545) in this dimension, which is also the highest score in all KRIs. Other indicators, including construction quality (D3.2, 4.273), technologies adopted in projects (D3.3, 3.909), and maintainability costs (D3.4, 3.818), all obtained high scores, which could indicate that maintainability received attention 04017046-8 J. Manage. Eng., 2018, 34(1): 04017046 J. Manage. Eng. Downloaded from ascelibrary.org by Tufts University on 10/28/17. Copyright ASCE. For personal use only; all rights reserved. from respondents because maintainability can help PPP projects improve the value and life span of facilities (Mirza 2006; Sobhiyah et al. 2009). Dimension of Operability As concluded from prior work, operability stressed both tangible and intangible assets (Yuan et al. 2015). According to the survey results, the indicators related to intangible assets are more important compared to the indicators related to tangible assets. The operation costs (D4.2, 4.091) and public service quality (D4.1, 3.955) are the most important indicators in this dimension, followed by project pricing (D4.5, 3.636), market share (D4.4, 3.591), and horizontal competition (D4.3, 3.591). The aforementioned indicators focus on measuring the operation management in PPPs to deliver highquality public service and meet the requirements of the general public. The incidents in maintenance (D4.6, 3.500) and technology improvement (D4.7, 3.409) are comparatively not important in this dimension, which indicates that operation depends more on the healthy running of a management system, as illustrated by Zhao et al. (2013). Dimension of Sustainability Li et al. (2005) indicated that sustainability had strong relationships with the success of PPP projects. The measures that help the public sector to achieve sustainability include the use of long-term contracts, output-based speciﬁcations, and ﬁnancial innovation. The survey results show that the reasonability of design (D5.1, 4.409) and construction quality (D5.2, 4.136) are the two most important indicators in this dimension, which can provide strong support to quality delivery of the PPP project to ensure that the drawings are available for maintenance referencing and the project is durable (Zhang 2006). Technology improvement (D5.3, 3.773) and environment inﬂuence (D5.6, 3.773) are both very important for sustainability of PPPs, which can facilitate guaranteed continued sufﬁciency (adequacy) of the project within its projected life span. Meanwhile, the comparatively high scores of maintenance costs (D5.4, 3.682) and public service quality (D5.5, 3.636) demonstrate that adequate ﬁnancial provision for maintenance and long-term services is critical. Dimension of Profitability and Refinanceability The indicators in this dimension can be classiﬁed into ﬁnancingrelated indicators and revenue-related indicators. For ﬁnancial indicators, the project’s self-liquidating ratio (D6.2, 4.182) and the availability of reﬁnancing (D6.3, 4.045) are the most important indicators and, concurrently, the most concerning indicators for private sectors (Xenidis and Angelides 2005; Appuhami et al. 2011). Financing structure (D6.1, 3.955) and costs (D6.4, 3.864) are determined by project planning, directly inﬂuencing availability of ﬁnancing and reﬁnancing of a PPP project (Wang et al. 2000; Wibowo et al. 2012), where the reﬁnanceability could be strongly inﬂuenced by the ﬂexibility of the initial ﬁnancing structure. Reﬁnancing would redesign the project’s initial ﬁnancial structure, utilizing an optimization potential by the contractor. With it, the concessionaire attempts to reduce the costs of the procurement of ﬁnancial services. For revenue-related indicators, constant or abundant cash ﬂow can impact the proﬁtability, which has strong relationships with operation costs (D6.7, 4.091), competition with a similar project directly invested by public sectors, market share (D6.8, 3.909), the price of the goods or services being offered (D6.10, 3.773), and maintenance costs (D6.11, 3.727). Obviously, operation costs, which reﬂect the management skills and experiences, are the most important. Additionally, the construction cost overrun (D6.6, 3.773) can also strongly inﬂuence the proﬁts of © ASCE private sectors, which could lead to penalties by the Chinese government in some PPP projects (Chan et al. 2009; Zhang et al. 2015). Construction delay (D6.5, 3.318) and technology improvements (D6.12, 3.318) are relatively not important in this dimension because cash ﬂow and earned money may be more signiﬁcant. Description of Survey Results Related to GKRIs The survey results of GKRIs are shown in Table 6. The ﬁndings revealed that political KRIs, including policy change (D7.1, 4.273) and local authority change (D7.2, 3.818), are the most signiﬁcant risks for PPP project success and RV improvement. The social KRIs cannot be ignored. The commitment and responsibility of both the public and private sectors are viewed as a signiﬁcant critical success factors (CSF) in PPPs (Li et al. 2005), which could also inﬂuence the payment of government and repayment from private sectors to loaners, reducing the RV of PPPs. Public opposition has occurred in many prior PPP projects in China, which heavily affected the success of PPP projects due to negative environment and social inﬂuence (Rouhani et al. 2016; Liu et al. 2017). In this case, the RV of PPPs would be inevitably reduced because the project could be terminated by public opposition. A legal risk comes about when new legislation and regulations are introduced with adverse consequences on existing transactions (Ke et al. 2010). The consequence of legal risks could particularly be a big problem for some transactions because the parties affected may not be able to perform their obligations (Iyer and Sagheer 2009). Therefore, the RV of PPPs obviously cannot be in accordance with the value prespeciﬁed in the PPP agreement. The economic KRIs were not important compared to political, social, legal, and systematic indicators. These indicators reﬂect the overall inﬂuences of external economic environment, and more obvious effects resulted from the internal indicators related to LKRIs in the dimension of proﬁtability and reﬁnanceability. However, the robustness of the ﬁnancial market is especially important in economic KRIs, which actually would inﬂuence the availability of ﬁnancing, ﬁnancing costs, ﬁnancing structure, and ﬁnancial viability (Salman et al. 2007). The systematic KRIs can be divided into indicators for an early stage and indicators for the whole process. For the indicators related to an early stage, the reasonability of decision making was the most important, which had close relationships with project planning, determination of design and ﬁnancing, concession period and pricing, and the issues that affect how the project is to be built. For the indicators related to the whole process, contract management is more important compared to force majeure and interface management. CFA-Based Refinement of KRIs for RVR in PPPs CFA on KRIs of RVR in PPPs Although the research survey indicated that the proposed KRIs were important, the large number of KRIs reduced the measurement efﬁciency for RVR evaluation. Therefore, reﬁnement of KRIs for RVR in PPPs is critical to improve the effectiveness and efﬁciency. The relationship between RVR and different KRI packages can be helpful to further reﬁne the KRIs for RVR in PPPs. Meanwhile, the mean values of different KRIs are also signiﬁcant for the reﬁnement process. A CFA was conducted by using the LISREL 8.70 software package to test whether the proposed conceptual model ﬁt the empirical data and further identify the important relationships between RVR and different KRI packages as well as between different KRI packages and KRIs. In CFA, the most common method used to estimate 04017046-9 J. Manage. Eng., 2018, 34(1): 04017046 J. Manage. Eng. Table 6. GKRIs for RVR in PPP Projects and Their Scores and Rankings Distribution shape Global risk dimension Political Social Legal Economic Downloaded from ascelibrary.org by Tufts University on 10/28/17. Copyright ASCE. For personal use only; all rights reserved. Systematic GKRIs Mean SD Ranking Skewness Skewness D7.1 Impacts of policy changes on RV D7.2 Impacts of changes of government ofﬁcials on RV D7.3 Impacts of creditworthiness for government on RV D7.4 Impacts of public opposition on RV D7.5 Impacts of legal changes on RV D7.6 Impacts of interest changes on RV D7.7 Impacts of exchange rate changes on RV D7.8 Impacts of inﬂation on RV D7.9 Impacts of imperfect ﬁnancing market on RV D7.10 Impacts of improper decision making on RV D7.11 Impacts of difﬁculty of land acquisition on RV D7.12 Impacts of difﬁculty of project approval on RV D7.13 Impacts of force majeure on RV D7.14 Impacts of contract management on RV D7.15 Impacts of interface management on RV 4.273 3.818 3.773 3.591 3.818 3.318 3.182 3.455 3.591 4.182 3.364 3.545 3.318 3.682 3.136 0.883 1.006 0.752 0.959 0.853 0.894 0.907 0.8 0.854 0.733 0.727 0.739 1.086 0.716 0.774 1 3 5 7 3 12 14 10 8 2 11 9 13 6 15 –0.054 –0.836 –0.327 –0.106 –0.637 –0.273 0.031 –0.451 –0.058 –0.304 0.115 –0.561 –0.219 0.569 –0.249 –0.054 –0.836 –0.327 –0.106 –0.637 –0.273 0.031 –0.451 –0.058 –0.304 0.115 –0.561 –0.219 0.569 –0.249 parameters is maximum likelihood (ML) estimation, which requires the assumption of multivariate normality of the observed data (Flora and Curran 2004; Li 2016). Researchers often use a Likerttype scale as a response format. However, the Likert scale data as a set of ordinal scales may violate the assumption. It would not be reasonable to expect that observed data would follow the normal distributional assumption, even in some important areas of research (Curran et al. 1996). In addition, the distribution shape of variables in Tables 5 and 6 indicates that the observed data basically accord with normal distribution, and the value of Cronbach’s alphas also veriﬁes reliability of the data. So the survey data in this study are suitable for performing CFA. Based on prior analytic research, CFA, as a special form of factor analysis and a conﬁrmatory technique, can be used to test whether the survey data ﬁt a hypothesized measurement model as presented in Fig. 3, including endogenous observed variables (RVR), exogenous latent variables (dimensions, i.e., D1–D7), exogenous observed variables (risk indicators, i.e., D1.1, D1.2,…, D7.15), errors in the variables, and pathway coefﬁcients (factor loadings). The latent variables that cannot be directly observed are measured with corresponding exogenous observed variables (indicators). The straight line from the latent variables (dimensions) to the corresponding observed variables indicates the causal effect on the observed variables (indicators). The factor loadings on the straight lines represent the relationship of indicators with their associated latent variables. The results of CFA demonstrated that the adequacy of the initial CFA model was good enough according to the model’s ﬁt to the data. A number of goodness-of-ﬁt indices of the initial model were used to evaluate how well the hypothesized model ﬁt the actual survey data. The results of commonly used ﬁt indices were as follows: normal theory weighted least-squares chi-square ( x 2) = 2,478.97, x 2/Df = 2.27 (Df, degrees of freedom = 1,091), comparative ﬁt index (CFI) = 0.95, root mean square error of approximation (RMSEA) = 0.022. Compared to the recommended values of these ﬁt indices as listed in Table 7 (Hu and Bentler 1999; Weston and Gore 2006), measured values of goodness-of-ﬁt indices indicated that the initial model of KRIs of RVR perfectly ﬁt the observed data. The errors in the variables and pathway coefﬁcients are shown in Fig. 3. According to the component of CFA framework, different KRI dimensions were all signiﬁcant (t > 1.96) and all contributed © ASCE differently to the change of RV in PPPs. The difference can be reﬂected based on the factor loadings and mean value of all KRIs in different dimensions. The dimension of maintainability obtained the highest loading (0.99), which indicates that maintenance is very important to the long-term running of PPP projects and could improve the value of PPP projects. As presented by Zhang (2006), maintainability is a necessary factor in the operation stage of PPP projects, which can improve the internal value of the facility and prolong the facility’s operation period. Meanwhile, the scores of sustainability (0.98) as well as proﬁtability and reﬁnanceability (0.95) are also high, which further demonstrates that these experienced respondents tended to focus on the future development and sustainable methods to explore a PPP project. The relatively low scores are from the dimensions of performance (0.77), function (0.44), operability (0.81), and GKRIs (0.76). Some KRIs in these four dimensions had comparatively higher loadings within corresponding dimensions for LKRIs. For the dimensions of performance and function, technical indicators were more important (e.g., design and quality). For the dimension of operability, management indicators are more important (e.g., costs and public services). In the dimension for GKRIs, there were higher loadings for the KRIs related to political changes, public opposition, legal change, and contract management in PPP projects. Refinement of KRIs for RVR in PPPs To further reﬁne the KRIs for RVR of PPP projects in China, the mean values obtained from survey analysis and factor loadings from CFA were both considered simultaneously. The KRIs, whose mean values were lower than 3.50 or factor loadings were lower than 0.400, are excluded from the initial model. Only the KRIs whose mean values were larger than 3.80 or factor loadings were larger than 0.600 are kept. For example, the mean value of D1.3, D1.4, and D1.6 is <3.50; in addition, the factor loadings of these three indicators are <0.400, therefore D1.3, D1.4, and D1.6 were excluded. The factor loading of D1.7 is also <0.400, but the mean value is >3.8, so D1.7 was kept. In this way, the excluded KRIs and kept KRIs are shown in Table 8, where 19 KRIs were excluded. Finally, 41 KRIs for RVR of PPPs were identiﬁed. In ﬁnal KRIs, many LKRIs should receive more attention. Fig. 4 shows the top-three indicators in each dimension. The designrelated indicators are the most important in the dimensions of performance, function, maintenance, and sustainability. In a PPP 04017046-10 J. Manage. Eng., 2018, 34(1): 04017046 J. Manage. Eng. Downloaded from ascelibrary.org by Tufts University on 10/28/17. Copyright ASCE. For personal use only; all rights reserved. Fig. 3. Loading estimates in CFA for proposed conceptual model project, design is usually the responsibility of private sectors, which would help private sectors reduce the lifecycle costs, improve the reliability of facilities, and greatly improve the residual value of PPP projects (Li et al. 2005; Clifton and Dufﬁeld 2006). Construction quality is the second most important for the RV of PPPs in the dimensions of performance, function, maintenance, and sustainability. High quality of infrastructure is delivered by the construction process, which will not only meet the stakeholders’ requirements, but also effectively improve the life span and durability of the project (Tang et al. 2013). Operation cost is the indicator for the operation stage, which is also critical for RV change in PPPs. Operation cost is the most important in the dimension of © ASCE operability, the second most important in the dimension of proﬁtability and reﬁnanceability, and ranks third in the dimension of function. The private sector has the responsibility for both the service and management of a facility to minimize performance deduction, proactively manage risks, and meet the operational requirements, in which operation cost control is especially important because the advantage of PPPs mainly depends on the lifecycle cost reduction and could be strongly inﬂuenced by operation costs (Love et al. 2015). In the operation stage, another important indicator is public service quality, which ranks second in the dimension of operability, fourth in the dimension of function, and ﬁfth in the dimension of sustainability. As innovative ways to deliver high- 04017046-11 J. Manage. Eng., 2018, 34(1): 04017046 J. Manage. Eng. Table 7. Common CFA Model Fit Indices and Their Acceptable Value Ranges Fit index x 2/Df (chi-square difference with 1 degree of freedom) CFI Downloaded from ascelibrary.org by Tufts University on 10/28/17. Copyright ASCE. For personal use only; all rights reserved. RMSEA Measured value Acceptable value range Acceptable values range from 1 to 5; <3 indicates good ﬁt A value of 0.90 or larger indicates a good model ﬁt; >0.95 indicates excellent ﬁt A value of 0.06 or less indicates an acceptable model ﬁt; below 0.05 indicates excellent ﬁt 2.27 0.95 0.022 quality public services, PPPs usually concentrate on value for money issues to change traditional public service delivery, allowing the private sector to introduce step change and cultural change, maximize synergies, generate internal investment, play a major role in meeting strategic and service objectives, and adopt output/ income-based measures for success (Sobhiyah et al. 2009). Another important ﬁnding from the research survey is that technology improvement and upgrade are viewed as key indicators for RVR of PPPs. In the dimensions of performance, maintenance, and sustainability, technology improvement and upgrade rank third in each dimension. Technical feasibility is the key to providing an imaginative technical solution for PPPs (Zhang 2006). Moreover, technology used in a PPP project is actually more signiﬁcant than in a traditional project. The technical details, advanced degree, update time of related equipment and devices, and time of technology upgrade should be considered carefully to improve the quality of resources and expertise available, design capability, management capability, commissioning ability, and adaptability of technology (Zhao et al. 2013; Osei-Kyei and Chan 2015). In ﬁnal KRIs, 12 GKRIs were kept, comprising political, social, legal, economic, and systematic indicators. The KRIs related to exchange rate, force majeure, and interface management were excluded, indicating that the investments from abroad are rare now in China, and internal management of special purpose vehicle (SPV) for PPP projects should strengthen the understanding of uncertainty of future changes. For political KRIs, the stability of local government (D7.2) was of great importance, indicating the creditworthiness of government is critical in improving RV (Abednego and Ogunlana 2006; Wibowo and Kochendoerfer 2011). Meanwhile, social KRIs (D7.3) were also very signiﬁcant as presented in prior researches, focusing on public satisfaction and social welfare of PPPs (Zhang 2011; Rouhani et al. 2016). Conﬁrmed by prior researches (Wang et al. 1999; Xu et al. 2010), legal changes greatly impact the RV changes. Furthermore, the degree of improvement in ﬁnancing market (D7.9) in China would also heavily inﬂuence the RV (Zhao et al. 2013; Chang and Chen 2016), which was the most signiﬁcant KRI from an economic perspective. Moreover, the perfect degree of contract management (D7.14) was always viewed as a crucial indicator for the success of PPP projects (Appuhami et al. 2011), which was the same in RVR management in PPPs from the viewpoint of systematic impacts. Potential Use of Identified KRIs In Fig. 3, the relationships between RVR, RVR dimension, and KRIs are clearly presented. To consolidate the number of KRIs, 21 © ASCE KRIs were removed according to the statistical analysis results, including factor loadings and statistical signiﬁcance. RVR management in PPP projects strongly depends on improvement of product or service performance, facility function, proﬁtability and possibility of reﬁnancing, maintainability, operability, and sustainability. Furthermore, political, social, legal, economic, and systematic impacts also strongly inﬂuence the outcome of PPP projects. The identiﬁed KRIs are useful tools for effective RVR management in PPPs. Potential use of these KRIs can be summarized as follows: • KRIs can be used to detect the weakness of RVR management during different stages of PPP projects. The identiﬁed six local risk factors and one global risk factor concurrently affect the RV during the whole concession period. The identiﬁed 41 KRIs can indicate how the risk factors inﬂuence the RV and detect the weakness of RVR management from seven dimensions. In the period of preconstruction, RVR management should pay high attention to the issues related to design and ﬁnancing. More emphasis of project design in PPP projects should be put on performance, function, maintainability, and sustainability because of the close relationship between RV and the impacts of design. Simultaneously, reasonable ﬁnancing analysis, structure, arrangement, and costs should be carefully implemented at an early stage of PPP projects. In the period of construction, time, cost, and quality control should be more important in inﬂuencing the RV, in which construction quality is the most important KRI that could impact the performance, function, maintainability, and sustainability. In the period of operation, maintainability and operation costs, market change, pricing change, technology progress, and horizontal competition are critical for the ﬁnal RV, in which maintainability and operation costs should be more signiﬁcant compared to other KRIs inﬂuencing all six local risk dimensions of RV. In the global risk dimension, policy change, government creditworthiness, legal change, decision-making ability, and contract management greatly inﬂuence the RV during whole concession period. • KRIs can be used to evaluate the RVR in PPPs. Six local risk factors and one global risk factor can be measured by different KRIs in different stages, and impact the RV of PPP projects during their life cycles. Possibility and degree of the loss of RV changes can be indicated by different KRIs and their criteria in different stages. Risk evaluations typically need corresponding data to calculate the possibility of RVR events and degree of RVR consequences according to different KRIs. Therefore, the appropriate KRIs must be determined to evaluate the RVR to estimate possible effects of any given change on the RV of PPP projects. Actually, the criteria of KRIs in various types of PPP projects could be different. When the KRIs are used in practical projects, more detailed criteria should be customized to the application. Furthermore, more data related to possibility of RVR events and degree of RVR consequences should also be collected through expert experiences and case studies to improve the accuracy of RVR evaluation. In contrast, the use frequency of KRIs to evaluate the RVR should be carefully considered. KRIs for RVR can be used throughout the concession period of a PPP project. However, ﬂexible arrangement of RVR management, including RVR evaluation frequency, should take the results of the risk evaluation and the requirements of government into account. If the results indicate low RVR level, frequent measurement may not be necessary because RVR evaluation demonstrates that the government would not suffer huge losses of RV when transferring the PPP projects. PPP projects would 04017046-12 J. Manage. Eng., 2018, 34(1): 04017046 J. Manage. Eng. Table 8. Reﬁnement of KRIs of RVR in PPPs Mean value <3.50 Downloaded from ascelibrary.org by Tufts University on 10/28/17. Copyright ASCE. For personal use only; all rights reserved. Dimension D1 D2 D1.3, D1.4, D1.6 D2.3, D2.4, D2.5, D2.6, D2.8, D2.9 D3 D4 D5 D6 D7 N/A D4.6, D4.7, D4.8 D5.4, D5.5 D6.5, D6.12 D7.6, D7.7, D7.8, D7.11, D7.13, D7.15 Factor loading <0.400 Kept KRIs Excluded KRIs D1.3, D1.4, D1.6, D1.7 D2.1, D2.2, D2.7, D2.8, D2.9 D1.7 (mean value >3.80) D2.1 (mean value >3.80) D2.2 (mean value >3.80) D2.3 (factor loading >0.600) D2.7 (mean value >3.80) N/A N/A N/A N/A D7.6 (factor loading >0.600) D7.8 (mean value >3.80) D7.11 ( actor loading >0.600) D1.3, D1.4, D1.6 D2.4, D2.5, D2.6, D2.8, D2.9 N/A D4.6, D4.7 N/A D6.10 D7.12 N/A D4.6, D4.7, D4.8 D5.4, D5.5 D6.5, D6.10, D6.12 D7.7, D7.12, D7.13, D7.15 Fig. 4. Top-three KRIs in different risk dimensions based on survey results suffer some loss of RV due to uncertainty from seven risk dimensions, and government may pay more attention to the RVR of projects during the concession period, suggesting a rationale for more frequent risk evaluation using KRIs. In general, RVR evaluation should be normalized work for government when monitoring the SPV of PPP projects. Conclusions This paper not only identiﬁes KRIs of RVR for PPP projects, but also tests the relationships between KRIs, risk dimensions, and RVR of PPP projects. A conceptual model of KRIs was proposed and further developed by analysis of the hypothesized relationships among indicators measuring PPP project RVR. A questionnaire survey was conducted to investigate opinions on seven risk dimensions and 61 KRIs that inﬂuence RVR of a PPP project before being transferred. Reliability analysis was conducted to test the internal consistency of the survey variable data, conﬁrming that opinions on KRIs are consistent, which indicate that the survey results can be further used to analyze the relationships among KRIs, risk dimensions, and RVR. The survey results demonstrated that most of the 61 KRIs are important (>3.00) and can be used to evaluate the © ASCE RVR from the seven risk dimensions. The CFA method was also used to test whether the hypothesized model correlates with data collected from the survey. The results of the CFA on the initial model showed a good model ﬁt, which indicated that all risk dimensions contribute to the change of RV in PPP projects, and that the classiﬁcation of KRIs within the risk dimensions is accurate (hypothesis in this paper). Based on the results of the survey and CFA, 61 KRIs were reﬁned to 41 KRIs, and the most important criteria for reducing RVR in PPP projects are effective maintenance, longterm sustainable development, and reasonable proﬁtability and reﬁnanceability. Additionally, active political supports, small social impacts, positive legal issues, stable ﬁnancing market, and proper decision making are also crucial to reduce RVR in PPPs. According to the statistical analysis and CFA, the greater value was assigned to high-quality design work in the early stages of PPP development, the lifecycle quality including construction quality and public service quality, comparatively low operation costs during the operation period, and feasible technologies adopted by private sectors. The 41 KRIs offer a useful tool for distinguishing strengths and weaknesses for effective RVR management and measurement in PPPs. Moreover, the CFA results provide a basis for reduced RVR and more effectively meet the requirement of improving the value of PPP projects to obtain long-term and sustainable development 04017046-13 J. Manage. Eng., 2018, 34(1): 04017046 J. Manage. Eng. Downloaded from ascelibrary.org by Tufts University on 10/28/17. Copyright ASCE. For personal use only; all rights reserved. through PPP projects. KRIs can indicate trends of losses and control weaknesses for RVR management in PPPs. A potential use of identiﬁed KRIs is to identify the weakness of RVR management and to measure the RVR in PPPs, where several different measurement methods can be adopted to monitor and calculate any speciﬁc RV changes in the conduct of PPP projects. Although this research on KRIs promotes improved understanding of RVR management in PPP projects, there are some limitations of the study. The causeand-effect relationships between the different risk dimensions should be clariﬁed through future research. The clariﬁcation of relationships between different KRIs will promote enhanced industry understanding of how to more effectively measure and reduce RVR of PPP projects, suggesting that further studies should explore how these KRIs can be most effectively applied. Acknowledgments The authors’ special thanks go to all reviewers of the paper and to the National Natural Science Foundation of China (NSFC71472037, 71671042); the Social Foundation of Jiangsu Province, China (13GLB005); the Program for Outstanding Young Teachers of Southeast University (2242015R30009); and the Fundamental Research Funds for the Central Universities for ﬁnancially supporting this research. References Abednego, M. P., and Ogunlana, S. O. (2006). “Good project governance for proper risk allocation in public-private partnerships in Indonesia.” Int. J. Project Manage., 24(7), 622–634. Akintoye, A., and Chinyio, E. (2005). “Private ﬁnance initiative in the healthcare sector: Trends and risk assessment.” Eng., Constr., Archit., Manage., 12(6), 601–616. Algarni, A. M., Arditi, D., and Polat, G. (2007). “Build-operate-transfer in infrastructure projects in the United States.” J. Constr. Eng. Manage., 10.1061/(ASCE)0733-9364(2007)133:10(728), 728–735. Ameyaw, E. E., and Chan, A. P. (2015a). “Evaluation and ranking of risk factors in public–private partnership water supply projects in developing countries using fuzzy synthetic evaluation approach.” Expert Syst. Appl., 42(12), 5102–5116. Ameyaw, E. E., and Chan, A. P. (2015b). “Risk ranking and analysis in PPP water supply infrastructure projects: An international survey of industry experts.” Facilities, 33(7/8), 428–453. Appuhami, R., Perera, S., and Perera, H. (2011). “Management controls in public–private partnerships- an analytical framework.” Aust. Accounting Rev., 21(1), 64–79. Arndt, R. H. (1998). “Risk allocation in the Melbourne City Link Project.” J. Structured Finance, 4(3), 11–24. Beasley, M. S., Branson, B. C., and Hancock, B. V. (2010) “How key risk indicators can sharpen focus on emerging risks.” Research Rep. of Committee of Sponsoring Organizations of the Treadway Commission hhttp://www.coso.org/documents/COSOKRIPaperFull -FINALforWebPostingDec110_000.pdfi (Dec. 19, 2016). Bonett, D. G., and Wright, T. A. (2015). “Cronbach’s alpha reliability: Interval estimation, hypothesis testing, and sample size planning.” J. Organiz. Behav., 36(1), 3–15. Brinkerhoff, D. W., and Brinkerhoff, J. M. (2011). “Public–private partnerships: Perspectives on purposes, publicness, and good governance.” Public Administration Dev., 31(1), 2–14. Bulsara, H. P., Kumar, A., Kumar, R., and Chauhan, K. A. (2015). “Experience of public private partnership in highway infrastructure development: An exploratory study of PPP mature countries and scenario in India.” Int. J. Procurement Manage., 8(5), 608–626. Burke, R., and Demirag, I. (2015). “Changing perceptions on PPP games: Demand risk in Irish roads.” Crit. Perspect. Accounting, 27, 189–208. © ASCE Burke, R., and Demirag, I. (2017). “Risk transfer and stakeholder relationships in public private partnerships.” Accounting Forum, 41(1), 28–43. Carbonara, N., Costantino, N., Gunnigan, L., and Pellegrino, R. (2015). “Risk management in motorway PPP projects: Empirical-based guidelines.” Transp. Rev., 35(2), 162–182. Chan, A. P. C., Lam, P. T. I., Chan, D. W. M., Cheung, E., and Ke, Y. (2009). “Drivers for adopting public private partnerships—Empirical comparison between China and Hong Kong special administrative region.” J. Constr. Eng. Manage., 10.1061/(ASCE)CO.1943-7862 .0000088, 484–494. Chan, A. P. C., Lam, P. T. I., Chan, D. W. M., Cheung, E., and Ke, Y. (2010). “Critical success factors for PPPs in infrastructure developments: Chinese perspective.” J. Constr. Eng. Manage., 10.1061 /(ASCE)CO.1943-7862.0000152, 1115–1124. Chan, A. P., Lam, P. T., Wen, Y., Ameyaw, E. E., Wang, S., and Ke, Y. (2015). “Cross-sectional analysis of critical risk factors for PPP water projects in China.” J. Infrastruct. Syst., 10.1061/(ASCE)IS.1943-555X .0000214, 04014031, Chan, A. P. C., Yeung, J. F. Y., Yu, C. C. P., Wang, S. Q., and Ke, Y. (2011). “Empirical study of risk assessment and allocation of public-private partnership projects in China.” J. Manage. Eng., 10.1061 /(ASCE)ME.1943-5479.0000049 136–148. Chang, C. Y., and Chen, S. (2016). “Transitional public–private partnership model in China: Contracting with little recourse to contracts.” J. Constr. Eng. Manage., 10.1061/(ASCE)CO.1943-7862.0001166, 05016011. Cheng, Z., Ke, Y., Lin, J., Yang, Z., and Cai, J. (2016). “Spatio-temporal dynamics of public private partnership projects in China.” Int. J. Project Manage., 34(7), 1242–1251. Cheung, E., and Chan, A. P. C. (2011). “Risk factors of public-private partnership projects in China: Comparison between the water, power, and transportation sectors.” J. Urban Plann. Dev., 10.1061/(ASCE)UP.1943 -5444.0000086, 409–415. Chou, J.-S., Tserng, H. P., Lin, C., and Yeh, C.-P. (2012). “Critical factors and risk allocation for PPP policy: Comparison between HSR and general infrastructure projects.” Transp. Policy, 22, 36–48. Clifton, C., and Dufﬁeld, C. F. (2006). “Improved PPP/PFI service outcomes through the integration of alliance principles.” Int. J. Project Manage., 24(7), 573–586. Comptroller and Auditor General of India. (2009). “Public private partnerships (PPP) in infrastructure projects: Public auditing guideline.” hhttp://infrastructure.gov.in/pdf/PPP-PROJECT.pdfi (Dec. 19, 2016). Curran, P. J., West, S. G., and Finch, J. F. (1996). “The robustness of test statistics to nonnormality and speciﬁcation error in conﬁrmatory factor analysis.” Psychol. Methods, 1(1), 16–29. De Los Ríos-Carmenado, I., Ortuño, M., and Rivera, M. (2016). “Private– public partnership as a tool to promote entrepreneurship for sustainable development: WWP torrearte experience.” Sustainability, 8(3), 199. Demirag, I., Khadaroo, I., Stapleton, P., and Stevenson, C. (2011). “Risks and the ﬁnancing of PPP: Perspectives from the ﬁnanciers.” Br. Accounting Rev., 43(4), 294–310. Dikmen, I., Birgonul, M. T., and Kiziltas, S. (2005). “Strategic use of quality function deployment (QFD) in the construction industry.” Build. Environ., 40(2), 245–255. Flora, D. B., and Curran, P. J. (2004). “An empirical evaluation of alternative methods of estimation for conﬁrmatory factor analysis with ordinal data.” Psychol. Methods, 9(4), 466–491. Fraser, J., and Simkins, B. J. (2011). Enterprise risk management: Today’s leading research and best practices for tomorrow executives, John Wiley & Sons, Hoboken, NJ. Frilet, M. (1997). “Some universal issues in BOT projects for public infrastructure.” Int. Constr. Law Rev., 14(4), 499–512. Froud, J. (2003). “The private ﬁnance initiative: Risk, uncertainty and the state.” Accounting, Organiz. Soc., 28(6), 567–589. Grimsey, D., and Lewis, M. K. (2005). “Are public private partnerships value for money?: Evaluating alternative approaches and comparing academic and practitioner views.” Accounting Forum, 29(4), 345–378. Hall, J. (1998). “Private opportunity, public beneﬁt?” Fiscal Studies, 19(2), 121–140. 04017046-14 J. Manage. Eng., 2018, 34(1): 04017046 J. Manage. Eng. Downloaded from ascelibrary.org by Tufts University on 10/28/17. Copyright ASCE. For personal use only; all rights reserved. Heckmann, I., Comes, T., and Nickel, S. (2015). “A critical review on supply chain risk—Deﬁnition, measure and modeling.” Omega, 52, 119–132. Hu, L., and Bentler, P. M. (1999). “Cutoff criteria for ﬁt indexes in covariance structure analysis: Conventional criteria versus new alternatives.” Struct. Eq. Model., 6(1), 1–55. Huang, X., Zhou, H., and Zhu, H. (2012). “Assessing the systemic risk of a heterogeneous portfolio of banks during the recent ﬁnancial crisis.” J. Financial Stab., 8(3), 193–205. Infrastructure Australia. (2008a). “National PPP guidelines Vol. 4: Public sector comparator guidance.” hhttp://www.infrastructureaustralia.gov.au /public_private/ﬁles/National_PPP_Guidelines-Vol_4_PSC_Guidance _Dec_08.pdfi (Dec. 19, 2016). Infrastructure Australia. (2008b). “National PPP guidelines Vol. 5: Discount rate methodology guidance.” hhttp://www.infrastructureaustralia.gov.au /public_private/ﬁles/National_PPP_Guidelines_Vol_5Discount_Rate _Methodology_Guidance_Jan_09.pdfi (Dec. 19, 2016). Iyer, K. C., and Sagheer, M. (2009). “Hierarchical structuring of PPP risks using interpretative structural modeling.” J. Constr. Eng. Manage., 10.1061/(ASCE)CO.1943-7862.0000127, 151–159. Jin, X. (2010). “Determinants of efﬁcient risk allocation in privately ﬁnanced public infrastructure projects in Australia.” J. Constr. Eng. Manage., 10.1061/(ASCE)CO.1943-7862.0000118, 138–150. Jin, X. H., and Zhang, G. (2011). “Modelling optimal risk allocation in PPP projects using artiﬁcial neural networks.” Int. J. Project Manage., 29(5), 591–603. Juan, C., Olmos, F., and Perez, E. (2016). “Decision support system to design feasible high-frequency motorways of the sea: A new perspective for public commitment.” Eng. Economist, 61(3), 1–27. Ke, Y., Wang, S., Chan, A. P., and Lam, P. T. (2010). “Preferred risk allocation in China’s public–private partnership (PPP) projects.” Int. J. Project Manage., 28(5), 482–492. Klijn, E. H., and Koppenjan, J. (2016). “The impact of contract characteristics on the performance of public–private partnerships (PPPs).” Public Money Manage., 36(6), 455–462. Koppenjan, J. F. M., and Enserink, B. (2009). “Public–private partnerships in urban infrastructures: Reconciling private sector participation and sustainability.” Public Admin. Rev., 69(2), 284–296. Krüger, N. A. (2012). “To kill a real option—Incomplete contracts, real options and PPP.” Transp. Res. Part A: Policy Pract., 46(8), 1359–1371. Lam, K. C., and Chow, W. S. (1999). “The signiﬁcance of ﬁnancial risks in BOT procurement.” Build. Res. Inf., 27(2), 84–95. Li, B., Akintoye, A., Edwards, P. J., and Hardcastle, C. (2005). “The allocation of risk in PPP/PFI construction projects in the UK.” Int. J. Project Manage., 23(1), 25–35. Li, C. H. (2016). “Conﬁrmatory factor analysis with ordinal data: Comparing robust maximum likelihood and diagonally weighted least squares.” Behav. Res. Methods, 48(3), 1–14. Li, J., and Zou, P. X. (2011). “Fuzzy AHP-based risk assessment methodology for PPP projects.” J. Constr. Eng. Manage., 10.1061/(ASCE)CO .1943-7862.0000362, 1205–1209. LISREL 8.70 [Computer Software]. Scientiﬁc Software International, Inc., Skokie, IL. Liu, J., Love, P. E., Sing, M. C., Smith, J., and Matthews, J. (2017). “PPP social infrastructure procurement: Examining the feasibility of a lifecycle performance measurement framework.” J. Infrastruct. Syst., 10.1061/(ASCE)IS.1943-555X.0000347, 04016041, Liu, T., Wang, Y., and Wilkinson, S. (2016). “Identifying critical factors affecting the effectiveness and efﬁciency of tendering processes in public–private partnerships (PPPs): A comparative analysis of Australia and China.” Int. J. Project Manage., 34(4), 701–716. Love, P. E., Liu, J., Matthews, J., Sing, C. P., and Smith, J. (2015). “Future prooﬁng PPPs: Life-cycle performance measurement and building information modelling.” Autom. Constr., 56, 26–35. Marques, R. C., and Berg, S. V. (2011). “Risks, contracts and private sector participation in infrastructure.” J. Constr. Eng. Manage., 10.1061 /(ASCE)CO.1943-7862.0000347, 925–932. © ASCE Mirza, S. (2006). “Durability and sustainability of infrastructure—A stateof-the-art report.” Can. J. Civ. Eng., 33(6), 639–649. Ng, S. T., Wong, Y. M. W., and Wong, J. M. W. (2010). “A structural equation model of feasibility evaluation and project success for public– private partnerships in Hong Kong.” IEEE Trans. Eng. Manage., 57(2), 310–322. Nunnally, J. C. (1978). Psychometric theory, 2nd Ed., McGraw-Hill, New York. Osei-Kyei, R., and Chan, A. P. (2015). “Review of studies on the critical success factors for public–private partnership (PPP) projects from 1990 to 2013.” Int. J. Project Manage., 33(6), 1335–1346. Partnerships Victoria. (2001). “Risk allocation and contractual issues.” hhttp://www.mfcr.cz/assets/en/media/PPP-Australia-Partnerships-Victoria -Risk-Allocation-and-Contractual-Issues.pdfi (Oct. 8, 2017). PMI (Project Management Institute). (2013). A guide to the project management body of knowledge (PMBoK® guide), 5th Ed., Newtown Square, PA. Private Finance Panel. (1996). Risk and reward in PFI contracts: Practical guidance on the sharing of risk and the structuring of PFI contracts, London. Ravindran, T. K. S. (2010). “Privatisation in reproductive health services in Pakistan: Three case studies.” Reprod. Health Matters, 18(36), 13–24. Rouhani, O. M., Geddes, R. R., Gao, H. O., and Bel, G. (2016). “Social welfare analysis of investment public–private partnership approaches for transportation projects.” Transp. Res. Part A: Policy Pract., 88, 86–103. Rwelamila, P. D., Fewings, P., and Henjewele, C. (2015). “Addressing the missing link in PPP projects: What constitutes the public?” J. Manage. Eng., 10.1061/(ASCE)ME.1943-5479.0000330, 04014085, Salman, A. F. M., Skibniewski, M. J., and Basha, I. (2007). “BOT viability model for large-scale infrastructure projects.” J. Constr. Eng. Manage., 10.1061/(ASCE)0733-9364(2007)133:1(50), 50–63. Seshadhri, G., and Topkar, V. (2016). “Validation of a questionnaire for objective evaluation of performance of built facilities.” J. Perform. Constr. Facil., 10.1061/(ASCE)CF.1943-5509.0000705, 04014191. Sharma, S. (2007). “Exploring best practices in public–private partnership (PPP) in e-government through select Asian case studies.” Int. Inf. Library Rev., 39(3–4), 203–210. Sobhiyah, M. H., Bemanian, M. R., and Kashtiban, Y. K. (2009). “Increasing VFM in PPP power station projects—Case study: Rudeshur gas turbine power station.” Int. J. Project Manage., 27(5), 512–521. Song, J. B., Song, D. R., Zhang, X. Q., and Sun, Y. (2013). “Risk identiﬁcation for PPP waste-to-energy incineration projects in China.” Energy Policy, 61, 953–962. SPSS 17.0 [Computer software]. IBM, North Castle, NY. State Council of China. (2014). “Opinions on strengthening the administration of local government debt.” hhttp://www.gov.cn/zhengce/content /2014-10/02/content_9111.htmi (Dec. 19, 2016). Tang, L., and Shen, Q. (2013). “Factors affecting effectiveness and efﬁciency of analyzing stakeholders' needs at the brieﬁng stage of public private partnership projects.” Int. J. Project Manage., 31(4), 513–521. Tang, L., Shen, Q., Skitmore, M., and Cheng, E. W. (2013). “Ranked critical factors in PPP brieﬁngs.” J. Manage. Eng., 10.1061/(ASCE)ME .1943-5479.0000131, 164–171. Thomas, A. V., Kalidindi, S. N., and Ganesh, L. S. (2006). “Modelling and assessment of critical risks in BOT road projects.” Constr. Manage. Econ., 24(4), 407–424. Wang, S. Q., Tiong, R. L. K., Ting, S. K., and Ashley, D. (1999). “Political risks: Analysis of key contract clauses in China’s BOT project.” J. Constr. Eng. Manage., 10.1061/(ASCE)0733-9364(1999)125:3(190), 190–197. Wang, S. Q., Tiong, R. L. K., Ting, S. K., and Ashley, D. (2000). “Evaluation and management of foreign exchange and revenue risks in China’s BOT projects.” Constr. Manage. Econ., 18(2), 197–207. Weston, R., and Gore Jr, P. A. (2006). “A brief guide to structural equation modeling.” Counsel. Psychologist, 34(5), 719–751. Wibowo, A., and Kochendoerfer, B. (2011). “Selecting BOT/PPP infrastructure projects for government guarantee portfolio under conditions of budget and risk in the Indonesian context.” J. Constr. Eng. Manage., 10.1061/(ASCE)CO.1943-7862.0000312, 512–522. 04017046-15 J. Manage. Eng., 2018, 34(1): 04017046 J. Manage. Eng. Downloaded from ascelibrary.org by Tufts University on 10/28/17. Copyright ASCE. For personal use only; all rights reserved. Wibowo, A., Permana, A., and Kochendörfer, B. (2012). “Modeling contingent liabilities arising from government guarantees in Indonesian BOT/ PPP toll roads.” J. Constr. Eng. Manage., 10.1061/(ASCE)CO.1943 -7862.0000555, 1403–1410. Witz, P., Leviäkangas, P., Łukasiewicz, A., and Szekeres, K. (2015). “Implementation of transport infrastructure PPPs in the Czech Republic, Finland, Poland and Slovakia—A comparative analysis on national contexts.” Int. J. Manage. Network Econ., 3(3), 220–237. Xenidis, Y., and Angelides, D. (2005). “The ﬁnancial risks in build–operate –transfer projects.” Constr. Manage. Econ., 23(4), 431–441. Xu, Y., Chan, A. P., Xia, B., Qian, Q. K., Liu, Y., and Peng, Y. (2015). “Critical risk factors affecting the implementation of PPP waste-toenergy projects in China.” Appl. Energy, 158, 403–411. Xu, Y. L., Sun, C., Skibniewski, M. J., Chan, A. P. C., Yeung, J. F. Y., and Cheng, H. (2012). “System Dynamics (SD)-based concession pricing model for PPP highway projects.” Int. J. Project Manage., 30(2), 240–251. Xu, Y. L., Yeung, J. F. Y., Chan, A. P. C., Chan, D. W. M., Wang, S. Q., and Ke, Y. (2010). “Developing a risk assessment model for PPP projects in China—A fuzzy synthetic evaluation approach.” Autom. Constr., 19(7), 929–943. Yuan, J., Chan, A. P., Xia, B., Skibniewski, M. J., Xiong, W., and Ji, C. (2016). “Cumulative effects on the change of residual value in PPP projects: A comparative case study.” J. Infrastruct. Syst., 10.1061/(ASCE)IS .1943-555X.0000272, 05015006. Yuan, J., Chan, A. P., Xiong, W., Skibniewski, M. J., and Li, Q. (2015). “Perception of residual value risk in public private partnership © ASCE projects: Critical review.” J. Manage. Eng., 10.1061/(ASCE)ME .1943-5479.0000256, 04014041. Yuan, J., Zeng, A. Y., Skibniewski, M. J., and Li, Q. (2009). “Selection of performance objectives and key performance indicators in public–private partnership projects to achieve value for money.” Constr. Manage. Econ., 27(3), 253–270. Yuan, J. F., Skibniewski, M. J., Li, Q. M., and Shan, J. (2010a). “The driving factors of China’s public-private partnership projects in Metropolitan transportation systems: Public sector’s viewpoint.” J. Civ. Eng. Manage., 16(1), 5–18. Yuan, J. F., Skibniewski, M. J., Li, Q. M., and Zheng, L. (2010b). “Performance objectives selection model in public-private partnership projects based on the perspective of stakeholders.” J. Manage. Eng., 10.1061/(ASCE)ME.1943-5479.0000011, 89–104. Zhang, X. (2011). “Social risks for international players in the construction market: A China study.” Habitat Int., 35(3), 514–519. Zhang, X. Q. (2006). “Public clients’ best value perspectives of public private partnerships in infrastructure development.” J. Constr. Eng. Manage., 10.1061/(ASCE)0733-9364(2006)132: 2(107), 107–114. Zhang, S., Gao, Y., Feng, Z., and Sun, W. (2015). “PPP application in infrastructure development in China: Institutional analysis and implications.” Int. J. Project Manage., 33(3), 497–509. Zhao, Z. Y., Zuo, J., and Zillante, G. (2013). “Factors inﬂuencing the success of BOT power plant projects in China: A review.” Renewable Sustainable Energy Rev., 22, 446–453. 04017046-16 J. Manage. Eng., 2018, 34(1): 04017046 J. Manage. Eng.