Mapping the Barriers of Utilizing Public Private Partnership into Brownfield Remediation Projects in the Public Land Ownership

Abstract

The financing issue is increasingly becoming a key problem for brownfield remediation in public land ownership, and Public Private Partnership (PPP) mode is considered a potentially effective solution. However, some barriers impede the utilization of the PPP mode into brownfield remediation projects in the situation of public land ownership. By taking China as an example, the study investigates the barriers when the PPP mode is used in brownfield remediation projects to deal with financing dilemmas. Specifically, 39 original barriers are first obtained from existing related literature. Based on these original barriers, a semi-structured questionnaire is designed and used in a Delphi process to achieve 14 final barriers, which can reflect the particular situation in China. To identify the interrelationship among these barriers, Interpretative Structural Modeling (ISM) is used to map the hierarchical structure of the final barriers, and the Cross-Impact Matrix Multiplication Applied to Classification (MICMAC) analysis is applied to show the relationship strengths of barriers. According to the results of the ISM and MICMAC analyses, three key barriers are determined, and several corresponding recommendations are provided from the perspective of the public administration.

1. Introduction

Along with the de-industrialization, a large number of factories have been moved out from cities, resulting in a large volume of unused or vacant construction land in urban areas, which is generally called “brownfield”. It has some particular characteristics: brownfield is the land that has been developed and used; part or all of the brownfield has been abandoned, idle, or underutilized; and brownfields may be polluted [1,2]. It is estimated that the number of brownfield sites in the United States is about 0.45 million [3], and China has over 2.62 million hectares of abandoned industrial land [4]. Brownfields usually have good locations and supporting infrastructures, which usually makes their redevelopment have higher economic benefits than the development of raw land on the edge of a city [5]. Moreover, brownfield redevelopment is also an effective way to prevent urban expansion [6] and deal with the shortage of urban construction land [7]. Therefore, many countries give high priority on their political agenda to brownfield redevelopment [8].

It is important to note that the intentions between “brownfield redevelopment” and “brownfield remediation” in China are different from that in other developed countries. In the land institution of private ownership, most initiate enterprises on brownfields are privately owned, and the pollution liability is relatively easy to determine. Therefore, the remediation of brownfields is generally a kind of private affair. Only when the responsible party of pollution is uncertain or unable to repair the brownfield, the government will take over the responsibility for repairing brownfields. Therefore, brownfield remediation and brownfield redevelopment are mutually inclusive and cannot be separated in most existing studies aimed at developed countries. While in China, land ownership belongs to the state or collective, and the public or organizations can only lease the use right of land for a certain period [9]. For urban construction land, the lease term ranges from 20 to 70 years. According to the land administrative law of China, most brownfields will be reacquired by the land reserve center of the local government. After finishing remediation, the new use right of brownfields will be traded at the primary land market by the way of “bid, auction, and listing”. Although the initiate enterprises are responsible to clear the pollution, many of them have gone through mergers and acquisitions, or even bankruptcy. Moreover, some key historical information on pollution is missing, and in turn the real responsible party is usually uncertain. For the land institution of public ownership in China, the local government is obligated to provide qualified construction land for the city and inevitably has to undertake the remediation liability of many brownfields [10]. For brownfields satisfying this condition, the remediation cost will have to be covered by the local finance, making brownfield remediation more treatable as public affairs. As a result, the remediation and redevelopment of brownfields in China are commonly separated.

Brownfield remediation is facing a series of challenges, such as immature remediation technology and long remediation cycle, and significant capital gap [11]. In China, the financing channels for brownfield remediation are quite few. The remediation expenditures generally come from enterprise funds, bank loans, and financial allocations from the local government. More seriously, many local governments in China are facing a budget deficit, which makes them unable to afford huge expenses [12]. Expanding financing channels for brownfield remediation in China is in urgent need. Some brownfield-related financial issues in developed countries have been discussed in existing literatures. The role of the government is emphasized [13,14,15,16,17]. Specifically, the government could effectively alleviate the funding gap of brownfield redevelopment projects through subsidies, tax relief, financial incentives, and even setting up specific funds. As a financing instrument for brownfield redevelopment, the effect of “Tax Increment Financing” was explored [18]. The PPP financing mode [19] was also proposed to improve the efficiency of brownfield-related projects. However, existing studies are mainly spotlighted on countries with private land ownership rather than public land ownership. Howland [20] examined three brownfield redevelopment cases in Baltimore to determine the boundary between the private and the public. Whitman [21] analyzed the market drivers and impediments of brownfield-related projects for investment from the perspective of private developers in Virginia. Glumac [22] explored the optimal PPP agreements of a brownfield redevelopment project for the municipality to attract private redevelopers in the Netherlands. Alexander [23] explored how brownfield redevelopment PPPs reflected network characteristics over time and the ways evolving network structures induced project outcomes in the state of New York. Li [24] discussed the compositions and characteristics of both public and private sectors of two brownfield redevelopment PPP projects in Pittsburgh. Studies on the financing issue of brownfields under the conditions of China are quite few. Han has proved that PPP will become a feasible financing mode for China’s brownfield remediation projects. What is more, potential key risks of brownfield remediation projects using PPP mode in China are determined by a systematic framework [25].

Although PPP is taken as a potential way to finance brownfield remediation projects by both scholars and officials in China, there is still no successful finished case in practice until now. PPP projects have features of high risk and long pay-off periods, the private investors are usually cautious to join them. Moreover, both theoretical research and practical experiences are lacking for utilizing the PPP mode into brownfield remediation projects in the land institution of public ownership, which will reduce the investment intention of the private. In summary, there are still multiple barriers to successfully applying the PPP mode to brownfield remediation projects. Only when these barriers are identified can they be dealt with one by one. This study attempts to find and map the barriers that prevent the utilization of PPP into brownfield remediation projects in China for supporting the local and central governments to improve related policy systems.

The contributions of this paper are threefold.

• The research enriches the financing issue of brownfield remediation projects in the context of public land ownership.
• It is the first time to systematically explore the barriers of adopting PPP into brownfield remediation projects in the land institution of public ownership.
• By using joint methods of ISM and MICMAC, three key barriers are determined and in turn some specific policies can be suggested.

The remainder of the paper is organized as follows. Section 2 introduces the methodologies of this study; Section 3 shows the identification and structure detection of the barriers; Section 4 explains the results; Section 5 discusses the additional insights and suggestions to the government. Finally, the conclusion is presented in Section 6.

2. Methods

The authors aim to use a two-step methodological framework to identify the barriers of PPP adoption in brownfield remediation projects in China. Firstly, the Delphi method is utilized in the barrier identification process. Secondly, the ISM is selected to present the structure of barriers, and the MICMAC is utilized to analyze the driving and dependence power of these barriers. Based on the results of ISM and MICMAC, the key barriers can be determined and furthermore support the local and central governments to make policies.

2.1. The Delphi Methods

The Delphi method, introduced by the RAND Corporation in 1946, is an effective survey method for collecting expert opinions. The Delphi method is a feedback anonymous method [26]. The general procedures of this method are as follows: obtaining opinions of experts and giving the result back to experts anonymously; asking for their opinions again and giving results back again until the consensus is obtained. The Delphi method has been applied to information systems [27], healthcare [28,29], and even brownfields [25,30] because of its advantages of anonymity, feedback, and convergence.

2.2. Interpretive Structural Modeling (ISM)

To get more insights into the interrelationship of the identified barriers, a hierarchical structure model needs to be utilized. ISM, developed by Professor Warfield [31] in 1974, is one of the most widely used approaches for structure analysis in the field of system engineering. The ISM methodology decomposes a complex, complete system into a structural model with multiple levels of progression, reflecting the different levels of importance of the components, in preparation for providing decision support to system managers. Compared with other widely used Multiple Criteria Decision Making (MCDM) methods, ISM has some specific advantages: (1) ISM categorizes a complex system into a multi-level recursive structural model for greater clarity; (2) ISM is widely used to analyze interactions among different factors/variables of a system; (3) ISM does not require any prior records of a system and prioritizes different factors according to their severity. ISM is suitable for this research because it is more macro-oriented and easy to execute [32]. The detailed steps to develop the hierarchical structure model of identified barriers by using the ISM method are as follows.

Step 1: The barrier set of utilizing PPP mode to finance brownfield remediation projects in China first needs to be determined.

Step 2: A structured self-interaction matrix model (SSIM) is constructed to examine the pair-wise relationship of barriers. The pair-wise comparison results are usually expressed by four letters, V, A, X, and O, and their detailed explanation is in Section 3.2.1.

Step 3: An adjacent matrix (AM) can be established based on the SSIM result. The AM is a “0–1” matrix, where “0” and “1” are used to replace the above-mentioned four letters. A detailed explanation is in Section 3.2.1.

Step 4: A reachability matrix (RM) can be obtained according to the transitivity rule, which is explained in Section 3.2.1 in detail.

Step 5: The barriers can be divided into several levels based on the results of RM through iterations.

Step 6: An ISM can be developed based on the above steps, and be presented as a hierarchical graph.

2.3. Impact Matrix Cross-Reference Multiplication Applied to a Classification (MICMAC)

Despite advantages, ISM also has limitations, as the model cannot describe the strength of the interrelationship among variables [33]. To make up for this disadvantage, the MICMAC analysis is carried out based on the principle of matrix multiplication. The driving and dependence power of each barrier can be calculated by the sum of the values of the row and column in the RM. The barriers can be classified into four clusters according to their driving and dependence power, namely: the autonomous cluster, the dependent cluster, the linkage cluster, and the driving cluster. The autonomous cluster includes barriers that have both weak driving and dependence power, which means they have almost no influence on other barriers. The dependent cluster includes barriers that have weak driving power but strong dependence power, which means that they are easily affected by other barriers and hardly affect others. The linkage cluster includes barriers that have both strong driving and dependence power, which means they can both influence or be influenced by other barriers in the system. The driving cluster includes barriers that have strong driving force but weak dependence, which means they often have an impact on other barriers and can be seldomly affected by others.

3. The Barrier Mapping Process of Utilizing PPP into Brownfield Remediation Projects

In this section, the authors aim to identify the barriers of utilizing PPP in brownfield remediation projects in China and map them to determine the key barriers. The initial barriers can be identified by literature review and the Delphi method. The ISM is utilized to clarify the interrelationship among barriers, and the MICMAC is utilized to analyze the interrelationship strength of these barriers. Based on the two methods, key barriers can be finally identified.

3.1. Identifying the Barriers

This study uses two steps to identify the barriers of utilizing PPP in brownfield remediation projects due to the rare relative research. The first step is to search for reasonable barriers from the existing literature. The second step is to use the Delphi method to optimize and finalize the barrier set.

3.1.1. Original Barriers from Literatures

A series of publications were examined by the authors. As a result, eight papers were pinpointed because of high citations, reputations, or recent publication time. Overall, 39 barriers have been identified and listed on the left of Table 1. The letters, “A to H”, stand for the 8 references, which are shown at the bottom of the table. For a specific barrier, check symbols (√) in the row indicate that this barrier is discussed in corresponding references.

Table 1. Original barriers from the literature.

No.	Barriers	Definition	References
			A	B	C	D	E	F	G	H
1	Absence of an enabling institutional environment	Lack of financial, legal, administrative, and other institutional environments to facilitate project implementation						√		√
2	Market demand change	Demand change from other factors, i.e., social, economic, etc., except the exclusive right		√	√
3	Site availability & preparation	Inadequate water and electricity supply, poor road access, and interference from surrounding residents may result in unavailability or interruption of use at some brownfield project sites			√	√			√	√
4	Poor political decision making	Government officials consider their career achievement or short-term goals or personal interests, or with little PPP experience, etc., resulting in a poor political decision-making process		√	√	√			√	√
5	High participation costs	The cost of private sector participation in the project is too high due to access barriers, etc.	√							√
6	Staging involvement of project participants	Project participants are subject to change due to long investment cycles and high uncertainty					√
7	Aversion to risk by project participants	Different government and private sectors have their risk appetite and will make decisions based on their risk appetite criteria					√
8	Excessive restriction on participation	Excessive restrictions on the entry threshold, background, and other factors for private sector participation in the project	√						√
9	Lack of political will	Government departments are not very motivated to promote this type of project due to various constraints						√	√
10	High risk relying on private sector	The private sector is limited in the amount of risk it can bear, the same risk affects the public sector differently than the private sector, and the potential exists for the private sector to take higher risks	√
11	Force majeure	Irresistible trends due to political situations, geological disasters, chemical contamination, etc., impede project implementation							√
12	Social distrust	Lack of trust in the surrounding residents and public opinion may generate adverse interference with the project				√			√
13	Confusion on government objectives and criteria evaluation	The government is bound by multiple forces and wants the project to achieve multiple goals and loses sight of the main objective	√
14	Poor financial market	There are too few financial institutions and financial products to provide financing for PPP projects			√	√			√	√
15	Thorough and realistic cost/benefit assessment	The project investment is huge, the cycle is long, and the return is not stable enough, so an accurate cost–benefit assessment is needed to ensure the safety of the project investment				√
16	Supporting utility risk	Supporting utilities, such as electricity and water, would not be available on time or at fair rates		√
17	Different sets of information about project risk	Project information elements, complex sources, different channels and capabilities of the participating parties to obtain information, which may produce information asymmetry					√		√	√
18	Environmental protection	Stringent regulation will have an impact on construction firms’ poor attention to environmental issues		√
19	Lack of trust among project participants	There is a trust gap between the government and private sectors					√		√
20	Complexity of contracts	To avoid risks, projects form very complex contracts that reduce the efficiency of execution					√		√
21	Immature juristic system	The lack of national PPP law leads to different ways of PPP implementation in different places in China		√		√			√	√
22	Project technical feasibility	Project execution requires complex technologies and the need to select the right ones				√				√
23	Government reliability	The reliability and creditworthiness of the government to be able and willing to honor their obligations in future		√	√	√			√
24	Competition	The government does not offer the exclusive right or does not honorits commitment and build another competitive project		√					√
25	Inadequate study and insufficient data	The lacking of relevant research makes the project implementation lacks reasonable references					√		√	√
26	Lack of experience or appropriate skills	Lack of relevant project successes and skills for reference	√						√	√
27	Government’s intervention	The public sector interferes unreasonably in privatized facilities/services		√		√
28	Competitive procurement process	The use of competitive procurement procedures is one of the main concerns of the private sector, which can increase uncertainty and costs for the private sector and reduce the willingness to participate				√			√	√
29	Low project value	Project value may become lower due to market changes, etc., reducing project revenue	√						√
30	The burden of local budget	Lack of sufficient funding from the government sector to support project operations							√	√
31	Weak Regulation framework	Immature legal environment, lack of relevant cases, resulting in a low level of supervision by regulators							√
32	Lack of Understanding benefits of optimal allocation	It is difficult for each department to hold a unanimous opinion on the distribution of benefits due to their vision and status					√
33	Lack of standard Model for PPP agreements	Lack of a standard model for PPP agreements to allocate risks and benefits					√		√	√
34	inappropriate risk Allocation and risk sharing	Partners are not able to identify and allocate risks well	√			√			√	√
35	Weak private consortium	The private sector’s financial strength is weak and not sufficient to bear the huge cost of project expenditures and long-cycle investments				√			√
36	Transparency in the procurement process	The more transparent the project procurement process is, the more it reduces transaction costs for the private sector and lowers the barrier to entry				√			√	√
37	High project costs	The actual cost of project implementation is too high	√						√	√
38	Corruption	Corrupt local government officials demand bribes or unjust rewards		√					√
39	Conflict beteewn local and federal government	Competition between local and central governments over funding and authority and responsibility								√

References: A = [25]; B = [34]; C = [35]; D = [36]; E = [37]; F = [38]; G = [39]; H = [40].

Table 1. Original barriers from the literature.

No.	Barriers	Definition	References
			A	B	C	D	E	F	G	H
1	Absence of an enabling institutional environment	Lack of financial, legal, administrative, and other institutional environments to facilitate project implementation						√		√
2	Market demand change	Demand change from other factors, i.e., social, economic, etc., except the exclusive right		√	√
3	Site availability & preparation	Inadequate water and electricity supply, poor road access, and interference from surrounding residents may result in unavailability or interruption of use at some brownfield project sites			√	√			√	√
4	Poor political decision making	Government officials consider their career achievement or short-term goals or personal interests, or with little PPP experience, etc., resulting in a poor political decision-making process		√	√	√			√	√
5	High participation costs	The cost of private sector participation in the project is too high due to access barriers, etc.	√							√
6	Staging involvement of project participants	Project participants are subject to change due to long investment cycles and high uncertainty					√
7	Aversion to risk by project participants	Different government and private sectors have their risk appetite and will make decisions based on their risk appetite criteria					√
8	Excessive restriction on participation	Excessive restrictions on the entry threshold, background, and other factors for private sector participation in the project	√						√
9	Lack of political will	Government departments are not very motivated to promote this type of project due to various constraints						√	√
10	High risk relying on private sector	The private sector is limited in the amount of risk it can bear, the same risk affects the public sector differently than the private sector, and the potential exists for the private sector to take higher risks	√
11	Force majeure	Irresistible trends due to political situations, geological disasters, chemical contamination, etc., impede project implementation							√
12	Social distrust	Lack of trust in the surrounding residents and public opinion may generate adverse interference with the project				√			√
13	Confusion on government objectives and criteria evaluation	The government is bound by multiple forces and wants the project to achieve multiple goals and loses sight of the main objective	√
14	Poor financial market	There are too few financial institutions and financial products to provide financing for PPP projects			√	√			√	√
15	Thorough and realistic cost/benefit assessment	The project investment is huge, the cycle is long, and the return is not stable enough, so an accurate cost–benefit assessment is needed to ensure the safety of the project investment				√
16	Supporting utility risk	Supporting utilities, such as electricity and water, would not be available on time or at fair rates		√
17	Different sets of information about project risk	Project information elements, complex sources, different channels and capabilities of the participating parties to obtain information, which may produce information asymmetry					√		√	√
18	Environmental protection	Stringent regulation will have an impact on construction firms’ poor attention to environmental issues		√
19	Lack of trust among project participants	There is a trust gap between the government and private sectors					√		√
20	Complexity of contracts	To avoid risks, projects form very complex contracts that reduce the efficiency of execution					√		√
21	Immature juristic system	The lack of national PPP law leads to different ways of PPP implementation in different places in China		√		√			√	√
22	Project technical feasibility	Project execution requires complex technologies and the need to select the right ones				√				√
23	Government reliability	The reliability and creditworthiness of the government to be able and willing to honor their obligations in future		√	√	√			√
24	Competition	The government does not offer the exclusive right or does not honorits commitment and build another competitive project		√					√
25	Inadequate study and insufficient data	The lacking of relevant research makes the project implementation lacks reasonable references					√		√	√
26	Lack of experience or appropriate skills	Lack of relevant project successes and skills for reference	√						√	√
27	Government’s intervention	The public sector interferes unreasonably in privatized facilities/services		√		√
28	Competitive procurement process	The use of competitive procurement procedures is one of the main concerns of the private sector, which can increase uncertainty and costs for the private sector and reduce the willingness to participate				√			√	√
29	Low project value	Project value may become lower due to market changes, etc., reducing project revenue	√						√
30	The burden of local budget	Lack of sufficient funding from the government sector to support project operations							√	√
31	Weak Regulation framework	Immature legal environment, lack of relevant cases, resulting in a low level of supervision by regulators							√
32	Lack of Understanding benefits of optimal allocation	It is difficult for each department to hold a unanimous opinion on the distribution of benefits due to their vision and status					√
33	Lack of standard Model for PPP agreements	Lack of a standard model for PPP agreements to allocate risks and benefits					√		√	√
34	inappropriate risk Allocation and risk sharing	Partners are not able to identify and allocate risks well	√			√			√	√
35	Weak private consortium	The private sector’s financial strength is weak and not sufficient to bear the huge cost of project expenditures and long-cycle investments				√			√
36	Transparency in the procurement process	The more transparent the project procurement process is, the more it reduces transaction costs for the private sector and lowers the barrier to entry				√			√	√
37	High project costs	The actual cost of project implementation is too high	√						√	√
38	Corruption	Corrupt local government officials demand bribes or unjust rewards		√					√
39	Conflict beteewn local and federal government	Competition between local and central governments over funding and authority and responsibility								√

References: A = [25]; B = [34]; C = [35]; D = [36]; E = [37]; F = [38]; G = [39]; H = [40].

3.1.2. Final Barriers from the Delphi Method

On one hand, the existing literature always has inappropriate points, such as not targeting brownfield remediation projects, or in the context of public land ownership. On the other hand, some feasible barriers may not be mentioned in the existing publications. Therefore, the original outcome needs to be screened and optimized. Firstly, a semi-structured questionnaire was designed based on Table 1. Secondly, experts need to be invited to give advice repetitively. Due to the lack of successful cases of adopting PPP in brownfield remediation projects in practice in China, it is difficult to invite experts directly involved in such projects. However, China has many experts involved in PPP projects or brownfield remediation projects. In this study, nine experts were invited to conduct the Delphi process. The experts include: (i) three professors from universities; (ii) two officials from land reserve agencies; (iii) two officials from land consolidation agencies; and (iv) two managers from land remediation companies.

In the first round, expert feedback collection by email took two weeks. Ten inappropriate original barriers were eliminated and one barrier was added to match the Chinese situation. The result is shown in the second column of

Table 2. Final barriers of PPP brownfield remediation projects in China.

No.	First Round	Second Round	Third Round (Final Barriers)
1	Inappropriate risk allocation and risk sharing	Inappropriate risk allocation and risk sharing	Unclear risk allocation (①)
2	Confusion on government objectives and criteria evaluation	Confusion on government objectives and criteria evaluation	Confusion on soil pollution background value (②)
3	Lack of experience or appropriate skills	Lack of experience or appropriate skills	Lack of experiences or skills (③)
4	High project costs	High project costs	Unclear benefit distribution (④)
5	Excessive restriction on participation	Excessive restriction on participation	Unstable project value (⑤)
6	High risk relying on private sector	High risk relying on private sector	Complicated decision making procedures (⑥)
7	Low project value	Low project value	Government reliability (⑦)
8	Poor political decision-making	Poor political decision-making	Environmental liability (⑧)
9	Government reliability	Government reliability	Poor financial market (⑨)
10	Supporting utility risk	Supporting utility risk & Site availability	Lack of scientific cost/benefit assessment (⑩)
11	Immature juristic system	Government’s intervention	Inadequate study and insufficient data (⑪)
12	Government’s intervention	Liability of environmental protection	Lack of trust between participants (⑫)
13	Liability of environmental protection	Poor financial market	Information asymmetry (⑬)
14	Poor financial market	Thorough and realistic cost/benefit assessment	Absence of an enabling institutional environment (⑭)
15	Site availability & preparation	Transparency in the procurement process
16	Project technical feasibility	Competitive procurement process
17	Thorough and realistic cost/benefit assessment	Inadequate study and insufficient data
18	Transparency in the procurement process	Aversion to risk by project participants
19	Competitive procurement process	Lack of understanding benefits of optimal allocation
20	Lack of standard model for PPP agreements	Lack of trust among project participants
21	Inadequate study and insufficient data	Different sets of information about project risk
22	Aversion to risk by project participants	Lack of political will
23	Lack of understanding benefits of optimal allocation	Absence of an enabling institutional environment
24	Lack of trust among project participants	High interest rate of private section
25	Complexity of contracts
26	Different sets of information about project risk
27	Lack of political will
28	Absence of an enabling institutional environment
29	Weak regulation framework
30	High interest rate of private section

. In the second round, expert feedback collection by email took one week. Another five inappropriate barriers were eliminated and two barriers were merged. The result is shown in the third column of Table 2. In the third round, expert feedback collection by email took two weeks. Nine barriers were eliminated, two barriers were merged because of their similar or overlap connotation, and one barrier was added. What is more, eight barriers made a better presentation. The result is shown in the fourth column of Table 2. In the fourth round, the experts did not give more different advances, and a consensus was reached. The final barriers in the fourth column of Table 2 were given new symbols with a circle. The detailed related documents for the Delphi process are shown in Supplementary Materials.

3.2. Identifying the Relationship among Barriers

ISM can transform the complex interrelationships among the barriers into an intuitive hierarchical framework, and MICMAC helps to further show the interrelationship strength of barriers.

3.2.1. ISM Analysis

The detailed procedures to develop the ISM model have been introduced in Section 2. Firstly, the barrier set including 14 barriers in Table 2 has been determined in Section 3.1. Secondly, the SSIM needs to be established according to the interrelationships among barriers. Four notations are used to explain the pair-wise comparison results of these barriers, and their definitions are explained as follows.

• V: barrier i influences barrier j, but j does not influence i;
• A: barrier i does not influence barrier j, but j influence i;
• X: barrier i and j influence each other;
• O: barrier i and j are unrelated.

Five Delphi experts from universities and land reserve agencies were further invited to judge the interrelationships among these barriers with the above four notations. The SSIM of the final barriers is shown in

Table 3. The SSIM result of barriers.

No.	①	②	③	④	⑤	⑥	⑦	⑧	⑨	⑩	⑪	⑫	⑬	⑭
①		A	A	V	O	V	A	V	V	X	O	V	A	A
②			A	O	O	O	O	V	O	O	O	O	O	O
③				O	O	O	O	O	O	V	A	O	O	A
④					O	O	A	O	O	A	O	V	O	O
⑤						O	O	A	O	O	O	O	O	O
⑥							O	O	O	O	O	O	O	A
⑦								O	O	O	O	V	V	O
⑧									O	O	O	O	O	O
⑨										X	O	O	O	A
⑩											O	O	A	O
⑪												O	O	O
⑫													A	O
⑬														O
⑭

. Due to the mapping property of the matrix, all of the comparison information has been contained in the upper right part of the matrix. For example, the letter V in row 1 and column 4 indicates that barrier ① influences ④, but not vice versa. Correspondingly, the letter in row 4 and column 1 would be judged as being A.

The third step is to build the AM of barriers. Based on the SSIM result in Table 3, the AM, a “0–1” matrix, can be developed. In this matrix, “0” and “1” are used to replace the above-mentioned four notations according to the following principles.

• If V is the value of barrier i to j in the SSIM, the (i, j) entry in the AM will be “1” and the (j, i) entry will be “0”.
• If A is the value of barrier i to j in the SSIM, the (i, j) entry in the AM will be “0” and the (j, i) entry will be “1”.
• If X is the value of barrier i to j in the SSIM, both the (i, j) and (j, i) entries will be “1”.
• If O is the value of barrier i to j in the SSIM, both the (i, j) and (j, i) entries will be “0”.

Following these transformation rules, the AM result of the barriers can be achieved and is shown in

Table 4. The AM result of barriers.

No.	①	②	③	④	⑤	⑥	⑦	⑧	⑨	⑩	⑪	⑫	⑬	⑭
①	0	0	0	1	0	1	0	1	1	1	0	1	0	0
②	1	0	0	0	0	0	0	1	0	0	0	0	0	0
③	1	1	0	0	0	0	0	0	0	1	0	0	0	0
④	0	0	0	0	0	0	0	0	0	0	0	1	0	0
⑤	0	0	0	0	0	0	0	0	0	0	0	0	0	0
⑥	0	0	0	0	0	0	0	0	0	0	0	0	0	0
⑦	1	0	0	1	0	0	0	0	0	0	0	1	1	0
⑧	0	0	0	0	1	0	0	0	0	0	0	0	0	0
⑨	0	0	0	0	0	0	0	0	0	1	0	0	0	0
⑩	1	0	0	1	0	1	0	0	1	0	0	0	0	0
⑪	0	0	1	0	0	0	0	0	0	0	0	0	0	0
⑫	0	0	0	0	0	0	0	0	0	0	0	0	0	0
⑬	1	0	0	0	0	0	0	0	0	1	0	1	0	0
⑭	1	0	1	0	0	1	0	0	1	0	0	0	0	0

. For example, the relationship notation of barrier ① and ④ is V. Based on the first principle, the value in the first row and fourth column should be “1” and that in the fourth row and first column should be “0”. Note that all values on the diagonal are set as being “0”.

In the fourth step, the AM result of the barriers can be converted into the RM result of the barriers in the light of the transitivity rule. The AM needs to be summed with a unit matrix, in which all the diagonal entries are assigned as being “1” because a given barrier influences itself. An example can be taken to explain the transitivity rule. Table 4 shows that barrier ① can influence barrier ④, and barrier ④ can influence barrier ⑫. Therefore, barrier ① can also influence barrier ⑫.

Table 5. The RM result of barriers.

No.	①	②	③	④	⑤	⑥	⑦	⑧	⑨	⑩	⑪	⑫	⑬	⑭	Driving
①	1	0	0	1	1	1	0	1	1	1	0	1	0	0	8
②	1	1	0	1	1	1	0	1	1	1	0	1	0	0	9
③	1	1	1	1	1	1	0	1	1	1	0	1	0	0	10
④	0	0	0	1	0	0	0	0	0	0	0	1	0	0	2
⑤	0	0	0	0	1	0	0	0	0	0	0	0	0	0	1
⑥	0	0	0	0	0	1	0	0	0	0	0	0	0	0	1
⑦	1	0	0	1	1	1	1	1	1	1	0	1	1	0	10
⑧	0	0	0	0	1	0	0	1	0	0	0	0	0	0	2
⑨	1	0	0	1	1	1	0	1	1	1	0	1	0	0	8
⑩	1	0	0	1	1	1	0	1	1	1	0	1	0	0	8
⑪	1	1	1	1	1	1	0	1	1	1	1	1	0	0	11
⑫	0	0	0	0	0	0	0	0	0	0	0	1	0	0	1
⑬	1	0	0	1	1	1	0	1	1	1	0	1	1	0	9
⑭	1	1	1	1	1	1	0	1	1	1	0	1	0	1	11
Dependence	9	4	3	10	11	10	1	10	9	9	1	11	2	1

shows the RM result of barriers, where the entry of (1, 12) has been “1”.

The fifth step is to divide the level of these barriers. From RM, each barrier corresponds to reachability and antecedent sets. The reachability set of barrier i (denoted as $R_i$) includes barriers that can influence others. $R_i$ contains the barriers with “1” in the i_th row of RM. The antecedent set of i (denoted as $Q_i$) includes barriers that are influenced by others. $Q_i$ contains the barriers with “1” in the i_th column of RM. For instance, $R_5$ include ⑤, $Q_5$ includes ①②③⑤⑤⑦⑧⑨⑩⑪⑬⑭.

Table 6. Reachability set, antecedent set and intersection.

No.	Reachability Set	Antecedent Set	Intersection	Level
①	① ④ ⑤ ⑥ ⑧ ⑨ ⑩ ⑫	① ② ③ ⑦ ⑨ ⑩ ⑪ ⑬ ⑭	① ⑨ ⑩
②	① ② ④ ⑤ ⑥ ⑧ ⑨ ⑩ ⑫	② ③ ⑪ ⑭	②
③	① ② ③ ④ ⑤ ⑥ ⑧ ⑨ ⑩ ⑫	③ ⑪ ⑭	③
④	④ ⑫	① ② ③ ④ ⑦ ⑨ ⑩ ⑪ ⑬ ⑭	④
⑤	⑤	① ② ③ ⑤ ⑦ ⑧ ⑨ ⑩ ⑪ ⑬ ⑭	⑤	1st
⑥	⑥	① ② ③ ⑥ ⑦ ⑨ ⑩ ⑪ ⑬ ⑭	⑥	1st
⑦	① ④ ⑤ ⑥ ⑦ ⑧ ⑨ ⑩ ⑫ ⑬	⑦	⑦
⑧	⑤ ⑧	① ② ③ ⑦ ⑧ ⑨ ⑩ ⑪ ⑬ ⑭	⑧
⑨	① ④ ⑤ ⑥ ⑧ ⑨ ⑩ ⑫	① ② ③ ⑦ ⑨ ⑩ ⑪ ⑬ ⑭	① ⑨ ⑩
⑩	① ④ ⑤ ⑥ ⑧ ⑨ ⑩ ⑫	① ② ③ ⑦ ⑨ ⑩ ⑪ ⑬ ⑭	① ⑨ ⑩
⑪	① ② ③ ④ ⑤ ⑥ ⑧ ⑨ ⑩ ⑪ ⑫	⑪	⑪
⑫	⑫	① ② ③ ④ ⑦ ⑨ ⑩ ⑪ ⑫ ⑬ ⑭	⑫	1st
⑬	① ④ ⑤ ⑥ ⑧ ⑨ ⑩ ⑫ ⑬	⑦ ⑬	⑬
⑭	① ② ③ ④ ⑤ ⑥ ⑧ ⑨ ⑩ ⑫ ⑭	⑭	⑭

summarizes the reachability set and antecedent set of each barrier in the second and third columns, separately.

The intersection ($R_i\cap Q_i$) of the two sets is shown in the fourth column of Table 6. Then, the intersection ($R_i\cap Q_i$) should be compared with the reachability set ($R_i$) for all barriers. If no differences exist, the corresponding barriers will form a level, and be removed from the barrier set. As the barrier in $R_5\cap Q_5$ is the same with that in $R_5$, barrier ⑤ is positioned to the first level and removed. Barrier ⑥ and ⑫ also follow this principle, and in turn are positioned to the first level too. Therefore, the first level contains barriers of ⑤ ⑥ and ⑫. The second level can also be determined according to this process. The repeating partition will be ended until all 14 barriers are included in a specific level. Ultimately, six levels are received and the partition result of the barriers is shown in

Table 7. The partition result of barriers.

Level	Barriers
I	⑤ ⑥ ⑫
II	④ ⑧
III	① ⑨ ⑩
IV	② ⑬
V	③ ⑦
VI	⑪ ⑭

.

3.2.2. MICMAC Analysis

The MICMAC method is applied to evaluate the driving and dependence power of the identified barriers, and then clarify the comprehensive influence of each barrier in the system. The concepts of driving and dependence power have been defined in Section 2.3. According to the gained RM in Section 3.1, the degree of the driving power of each barrier is determined by the sum of “1” appearing in the corresponding row. The degree of dependency power is determined by the sum of “1” in the corresponding column. The results of the driving and dependence power for each barrier are shown in the last row and last column of Table 5.

4. The Barrier Mapping Results of Utilizing PPP into Brownfield Remediation Projects

4.1. The Hierarchical Framework of the Final Barriers

According to the RM in Table 5 and the partition result in Table 7, a visual hierarchical framework of the final barriers is shown in Figure 1. All barriers are allocated at the corresponding levels and the arrow points to the influenced barrier by other connected barriers.

The barriers at the bottom level are “inadequate study and insufficient data” (⑪) and “absence of an enabling institutional environment” (⑭), which are the basic barriers with a fundamental and deep impact. They need to be resolved as high priority. The barriers at the top level include “unstable project value” (⑤), “complicated decision making procedures” (⑥), and “lack of trust between participants” (⑫), which are the ultimate influence target of the system. Most of these problems need to be solved through the bottom and intermediate layer barriers. The barriers in the intermediate layer are “lack of experiences or skills” (③), “unclear benefit distribution” (④), “confusion on soil pollution background value” (②), “government reliability” (⑦), “information asymmetry” (⑬), “unclear barriers allocation” (①), “environmental liability” (⑧), “poor financial market” (⑨), and “lack of scientific cost-benefit assessment” (⑩).

4.2. The Classification of the Final Barriers

A visual MICMAC matrix result of the barriers is shown in Figure 2. As can be seen, there are four clusters in it. Cluster-I is the autonomous cluster that has weak dependence and driving force, and there are no barriers in Cluster-I. Cluster-II is the dependent cluster that has strong dependence power but weak driving force, and it covers five barriers. Cluster-III is the linkage cluster that possesses both strong driving and dependence power, and it contains three barriers. Cluster-IV is the driving cluster that has strong driving power but weak dependence power, and six barriers are involved in Cluster-IV. As can be seen, barriers in Cluster-III and Cluster-IV have relatively high driving power, which indicates they have more influence on others.

5. Discussions

In countries of public land ownership, the government has relatively more power and resources. When these power and resources encounter barriers in promoting public utilities, cooperating with the private sector to implement PPP projects is a very effective way to solve financing difficulties and improve operational efficiency. The research on the application barriers of PPP in brownfield remediation projects has broadened its application in China and other public ownership countries. Moreover, further suggestions are provided to support the Chinese government to arrange the utilization of PPP in the brownfield remediation area.

5.1. Additional Insights

Examining Figure 1, all barriers can be sorted into three classes: source, process, and result barriers. The source barriers (⑦⑪⑭), marked in grey, are those that can only influence other barriers whereas not vice versa. These barriers have two characteristics: Firstly, it is a reality formed by a complex background that could not be changed easily; Secondly, they are stable and inevitable, and are determined by the nature of affairs. The result barriers can only be influenced by other barriers, but not the opposite, such as ⑤⑥⑫. The process barriers, with arrows entering and leaving, can either influence or be influenced by other barriers. From the result of ISM, source barriers are the most important barriers because other barriers are directly or indirectly influenced by them. According to MICMAC analysis, the barriers in Cluster-III and Cluster-IV have a high influence on other barriers, which indicates that they are critical in the system. Based on the results of ISM and MICMAC, three barriers are taken as being the key barriers: “government reliability” (⑦), “inadequate study and insufficient data” (⑪), and “absence of an enabling institutional environment” (⑭). The result is consistent with the Pareto optimality principle: 20% of causes may induce 80% of results.

Considering the “government reliability” (⑦), the political system of China makes the government have the function of “operator”. In PPP projects, the government needs to change its role from “operator” to “supervisor”, and the government tends to over intervene in PPP projects, which is caused by the unequal position of the government and social capital in the market. Although the government commitment is attractive, there is still a high risk of violating the commitment for the investors. Moreover, it is difficult for the private to obtain reasonable compensation due to the unequal power. Considering the “inadequate study and insufficient data” (⑪). The lack of relevant studies has been fully discussed earlier, and the absence of cases and studies has contributed to the lack of data. This barrier makes the private party worried about the risk and profit of investing in brownfield remediation projects in China because no successful case can be referenced in practice. Considering the “absence of an enabling institutional environment” (⑭), there is no special law for the implementation of PPP mode in brownfield remediation projects in China. The ministerial regulations, local government management regulations, and franchise management methods formulated for specific projects have insufficient authority and poor stability. The dispute settlement mechanism is relatively unsound.

5.2. Suggestions to the Government

Several policies are suggested to help overcome the three key barriers of the utilization of PPP in brownfield remediation projects in China.

(1) Enhancing government reliability

Sometimes, the public party may use the power advantage to make decisions that are more beneficial for themselves. This leads to the private party’s concern about government reliability. A restraint mechanism of local government power needs to be formed to balance the power asymmetry between the public and the private. The central government can explore establishing a default guarantee fund to guarantee the private sector be able to compensation directly from the fund instead of having to resort to complex litigation. Moreover, a set of credit rating mechanisms for the local government needs to be established to assess their reliability. Based on it, the private sector can make decisions and avoid risks.

(2) Setting up pilot projects and conducting basic research

Several pilot projects should be set up to collect experiences, so as to reduce investors’ fears about the failure of the brownfield remediation PPP projects. This could also lay the foundation to establish a contract template for subsequent similar projects. Some basic research should be conducted to make some critical issue clearer, such as the suitable risk allocation scheme, contract type, and responsibilities of each party [41].

(3) Improving the institutional environment

It is necessary to improve the brownfield-related law system. The potential liability for land contamination will increase the risk uncertainty, which may decrease the attractiveness of brownfield remediation projects for the private sector. Some developed countries have issued similar legislation for China to be referenced, such as the Landfill Tax Credit Scheme in the United Kingdom [42] and Pact for a Green New Deal in Canada [43]. Moreover, the financial market in China should be improved. In the public land ownership country, the power between the private and the public is unequal in the financial market. More financing instruments are needed to help the private sector be able to finance brownfield remediation projects. Suitable brownfield-related insurance to control risk is also in urgent need by the private sector. In developed countries like the US and Australia [44], the financial system for brownfield remediation is well-developed, and China can learn from it to some extent.

6. Conclusions

The PPP is a potential way to solve the financing problems of brownfield remediation projects in public land ownership like China. However, there is still no successful finished brownfield remediation PPP projects in practice in China until now. Systematical studies on the barriers of utilizing PPP into brownfield remediation projects are absent. This research tried to explore this issue in the background of China. Specifically, 39 relevant barriers were first obtained from the existing literature. They were combined, revised, and expanded by using the Delphi method, resulting in a list of 14 final barriers. Furthermore, a 6-level ISM was established to figure out the interrelationship and hierarchy of these barriers. Based on the ISM result, the MICMAC analysis method was used to analyze the driving and dependency power of the 14 final barriers. Cluster-I to Cluster-IV in the MICMAC matrix contains zero, five, three, and six barriers, respectively. Considering the intersection of the results of the ISM and MICMAC analyses, three barriers were identified as key factors: “government reliability”, “inadequate study and insufficient data”, and “absence of an enabling institutional environment”. Specific attention should be given to these key barriers, and in turn some practical implications were proposed for the Chinese government to improve the current situation.

The findings of this paper can help the government understand the barriers to using PPP into brownfield remediation projects and support a more effective decision-making process. Future research should be conducted aiming at the identified key barriers so as to increase the attraction of brownfield remediation projects for the private sector.