Public-Private Partnership Transportation Investment and Low-Carbon Economic Development: An Empirical Study Based on Spatial Spillover and Project Characteristics in China

Abstract

As the most widely used investment mode of transportation infrastructure in China, PPP has been regarded as an effective institutional tool by the Chinese government for the construction of efficient and sustainable transportation infrastructure, promoting high-quality economic development. In order to evaluate the economic sustainability of PPP, this paper innovatively takes the low-carbon economy as the research perspective and uses 276 cities of China from 2006 to 2018 as samples to comprehensively investigate the impact of transportation PPP investment on the low-carbon economy from the diversified dimensions of the overall effect, spatial effect, mediating effect, heterogeneous effect, lagged effect, and PPP project-level characteristics. The results show that: (1) transportation PPP investment markedly promotes low-carbon economic development in China, which is manifested as the dual improvement effects of benefits and efficiency; (2) transportation PPP investment has a positive spatial spillover effect that can promote the low-carbon economic development of neighboring cities; (3) transportation PPP investment can promote low-carbon economic development by promoting the upgrading of industrial structures, and furthermore has a spatial transmission mechanism; (4) impacts of transportation PPP investment in different subsectors are significantly heterogeneous; (5) transportation PPP investment has a continuous and stable impact on promoting the low-carbon economy in China; (6) the participation of state-owned enterprises and listed companies in PPP projects has a significant positive impact, and state-owned capital is more effective in projects with relatively large scale. All these provide reliable evidence for the far-reaching significance and practical value of transportation PPP investment on sustainable low-carbon economic development.

1. Introduction

In the new age, the low-carbon economy has become an important development trend worldwide, which reveals the fundamental transformation of the human survival concept and economic development mode. As a major carbon emitter, China puts forward the environmental targets of carbon peaking and carbon neutrality, and the sustainability strategy of high-quality economic development. Transportation is a pioneer in the economic development of China, which has an priority mission of sustainable development. In the face of the increasing demand of transportation infrastructure construction, the supply mode which takes the government as the single body of investment and operation gradually exposes the drawbacks of the low supply efficiency and huge financial burden. At this point, a Public-Private Partnership (PPP) mode is highly promoted by the Chinese government due to its efficiency advantages. According to China’s Ministry of Finance, by June 2022, China’s total PPP investment in transportation had reached 5.8 trillion yuan, making China the world’s largest transportation PPP market. The 14th Five-Year Plan of China emphasizes giving full play to the role of PPP to sustain the efficient construction of green infrastructure and major transportation infrastructure. Experts from the Ministry of Finance said that PPP mode has become a new driving force for the low-carbon economy with its capital, management, and technological advantages. In this context, PPP has great potential under the development goals of the low-carbon economy.

Most studies have confirmed the economic value of PPP and point out that PPP has economic growth effect, not only in China [1,2,3], but also in other countries [4,5,6]. On the indirect side of the economic development, PPP can promote the quality and efficiency of infrastructure construction [7], improve employment and social welfare [8], drive the upgrading of industrial structure [9], and achieve higher-level urbanization [10]. In terms of spatial research, Cai’s empirical study [6] found that PPP had a significant positive spatial spillover effect on the economic growth of the Belt and Road countries. Chen [11] confirmed the positive spatial spillover effect of PPP on regional economic development in China. By comparing the economic consequences of PPP with the traditional mode, some studies found that PPP applied in transportation infrastructure had more positive economic impacts than other fields [12], and PPP had a stronger promotion effect on transportation infrastructure investment efficiency than single government investment [13]. Chotia and Rao [5] compared the impact of PPP and single investment mode on economic development in India and found that only the PPP mode had significant economic growth effect. About PPP and carbon emissions, some studies pointed out that transportation PPP investment had certain positive effects on reducing carbon emissions from transportation industry [14]; and PPP could play an important role in low-carbon development by supporting green infrastructure [15]. However, Yao et al. [16] and Pei et al. [17] indicated that PPP was not a “panacea”, and it could bring more efficiency only when certain conditions are met. To sum up, few studies have considered the impact of the PPP mode on economic growth from the perspective of low-carbon development. Transportation PPP investment links the transportation industry, ESG investment, and social cooperation in one, which embodies diversified economic attributes. According to the existing findings, the efficiency advantages, economic growth effects of PPP, and its role in low-carbon development imply that it may play an active role in the low-carbon economy. Based on the green sustainable development concept and trend, it is necessary and worthy to connect carbon emissions with economic growth and investigate the impact of transportation PPP investment on the low-carbon economy.

On these grounds, this paper uses hand-collected, unique data of 276 cities in China from 2006 to 2018 as samples, based on the evaluation of the level and efficiency of low-carbon economic development in China, and empirically examines the impact of transportation PPP investment on China’s low-carbon economic development from the diversified dimensions of the overall effect, spatial spillover effect, heterogeneous effect, mediating effect, lagged effect, and PPP projects characteristics. The main contributions of this study include revealing the relationship between transportation PPP investment and the low-carbon economy for the first time, which enriches the existing research on the economic consequences of PPP and confirms its profound influence and practical value; this provides empirical evidence and a point of reference for encouraging the widespread application of PPP in transportation infrastructure so that it may play a more effective and sustainable role in low-carbon economic development. For other countries, this paper may provide some valuable points of reference and a basis for them to apply or innovate PPP modes in the field of transportation infrastructure or other fields seeking to support low-carbon economic development.

The paper is organized as follows: the theoretical analysis is presented in Section 2. The data, variables, and model are introduced in Section 3. Section 4 presents the analyses of the evaluation of China’s low-carbon economy, and the empirical analyses of the impact of transportation PPP investment on the low-carbon economy. Section 5 explains and discusses the findings. Section 6 presents the conclusion and enlightenments.

2. Theoretical Analysis

2.1. Transportation PPP Investment and Low-Carbon Economic Development

The low-carbon economy is a comprehensive and multi-objectives system including the multivariance of economic, ecological, social, management, and institutional elements with the goal of realizing higher economic growth with lower carbon emissions, which is driven by various factors of industrial upgrading, technological innovation, institutional change, resource allocation, effective investment, and so on. Therefore, all factors that can affect economic growth and carbon emissions will affect the low-carbon economy. Firstly, through the investment multiplier effect, PPP can stimulate domestic demand and expand the effective supply of transportation infrastructure to consolidate the endogenous foundation for low-carbon economic development [9]. Secondly, by the connectivity of transportation infrastructure under massive PPP investment, the high-end production factors such as management, technology, and knowledge can be allocated in a wider range of contexts. Furthermore, PPP introduces a marketization mechanism to optimize resource allocation, whose application in transportation infrastructure can conspicuously reduce the flow and use costs of production factors and improve resource allocation efficiency [18] to reduce the resources required by unit economic output in the process of economic development [19] thus forming the overall effect of energy conservation and emissions reduction throughout society [20]. Moreover, through PPP, large-scale construction of all kinds of transportation infrastructure can provide an indispensable carrier foundation for the upgrading of industrial structures, the progress of new urbanization and technological innovation of enterprises [10,21], thus promoting the improvement of various links in the low-carbon economy. The PPP mode can not only introduce abundant funds from social capital, but also simultaneously introduce their experience, technology, and knowledge [22] to achieve higher quality operations of transportation infrastructure, thus contributing to the construction of green transportation systems. Besides, under the mutual restriction of government and social capital, the downsides of government corruption and profit-seeking by social capital have also been alleviated by the governance mechanism of PPP [23,24], which will promote the sustainability and coordination between social benefits and economic benefits.

2.2. Spatial Spillover Effect of Transportation PPP Investment

Transportation infrastructure has the characteristics of networking and externality. It can break the spatial and temporal distance between each geographical space and connect it into a complete whole, thus speeding up the free flow of economic elements among regions and further changing the regional population distribution, industrial agglomeration, and economic space of cities. Moreover, transportation infrastructure has a strong effect of gathering elements and resources around and along its routes, thus making economic development present the spatial characteristics of agglomeration and scale. Therefore, transport PPP investment, as an important source of transport infrastructure construction, generates spatial spillover effect through expanding the supply scale, improving the spatial layout, and optimizing the structure of transportation infrastructure. According to economic theory, the externality of infrastructure roots in its public goods attribute, which can be further divided into positive externalities and negative externalities. The stronger externalities of transportation infrastructure compared to other forms of infrastructure have been widely supported by studies but whether it is positive or negative has not been unified. Zhang [25] pointed out that transportation infrastructure had an overall positive external effect on regional economic growth in China, but there was also evidence to show a negative or inapparent effect [26]. Furthermore, there are differences in the externalities of different types of transportation infrastructure [27]. From the perspective of PPP mechanism, transportation infrastructure invested through PPP is more efficient [7,28] and can bring a greater positive economic impact than traditional modes [12]. Therefore, it is reasonable to consider that transportation PPP investment can produce positive spatial spillover effect in general.

2.3. Mediating Effect of Industrial Structure on PPP and the Low-Carbon Economy

In the current process of industrialization in China, the unreasonable industrial structure and the development mode of high input, high pollution, and low output of the secondary industry will have adverse effects on the sustainability of a low-carbon economy. The transformation of the industrial structure, especially the decline of the proportion of the secondary industry, is an effective way to reduce carbon emissions and promote economic sustainability [29]. As a result, accelerating the dominance of the tertiary industry in GDP has become China’s long-term development vision. Transportation infrastructure is a vital carrier for the development of the modern industrial system, its efficient construction and operation construction by PPP mode is a significant source for ensuring large-scale supply of transportation infrastructure and optimizing regional transportation connectivity, which can improve the spatial flow and reasonable allocation of production factors among different industrial sectors so as to promote the upgrading of regional industrial structures [9]. Additionally, the tertiary industry can benefit more from the improvement of transport infrastructure due to more convenient and less costly exchange of advanced production factors such as knowledge and technology [30]. On these grounds, this paper considers that there is a chain relationship between transportation PPP investment and the low-carbon economy, that is, PPP can promote the upgrading of industrial structure, thus further driving low-carbon economic development.

2.4. Heterogeneity of Transportation PPP Investment in Different Subsectors

Different transportation infrastructures have different investment attributes and economic consequences. Most studies have shown that railway and airport infrastructure play a stronger role in driving regional economic integration through its higher transportation efficiency and more complete network system [31,32]. As the basis for connectivity and accessibility of transportation network and the field with the largest number of PPP projects, road infrastructure has significant economic growth effect and positive spatial spillover effect [33], while its spatial spillover effect is weaker than that of railways [34]. Additionally, a subway is an intra-city infrastructure which has positive externality, manifesting as the value-added effect of commerce and land along the line to promote urban economic development [35,36]. Furthermore, PPP investment in different transportation subsectors will also generate different effects on resource allocation, industrial transformation, and economic agglomeration, which are highly correlated with the differentiated nature, function, and sphere of transportation infrastructure.

2.5. Lagged Effect of Transportation PPP Investment

The lagged effect reveals the time dynamic correlation between variables and is an important aspect of judging sustainability. Transportation PPP investment is a long-term investment, which is reflected in the long-term attribute of return on investment, the long franchise period of PPP projects, and lasting functionality of transportation infrastructure. Due to the inertia of economic activities, the connection between economic indicators tends to continue and delay, thus forming a hysteresis in time. These determine the lagged effect of transportation PPP investment on the low-carbon economy, which conforms to economic law. According to the influencing mechanism of PPP on the low-carbon economy, in addition to the immediate impact, PPP will indirectly affect many economic elements including the industrial structure, technological innovation, resource allocation, and so on; these effects are not apparent in the short term. Furthermore, the long-term operation of transportation infrastructure will continuously accumulate strength for economic development, which also takes time to prove. Consequently, it will be considered whether transportation PPP investment can play a stable and sustaining role in promoting the low-carbon economy.

2.6. Impact of Characteristic Factors of Transportation PPP Projects

In the institutional context of China, PPP has typical characteristics of the “Chinese style”, the most striking aspect of which is that the social capital participating in PPP projects is mainly state-owned enterprises [37]. State-owned enterprises have significant advantages in economic strength, technological innovation, and management ability, and are more resilient to risk which meets the demands of large-scale construction of transportation infrastructure and contributes more to the success of the PPP projects. Additionally, driven by the ESG investment idea, more and more listed companies take the initiative to participate in PPP projects, especially for transportation infrastructure projects related to the national economy and people’s livelihood. The participation of listed companies is not only the result of high support from the Chinese government, but also the vision of listed companies to fulfill social responsibilities and establish a good reputation, which has also triggered a positive market reaction and boosted their share prices [38]. This stimulates listed companies to make efforts to improve the performance of PPP projects. In addition, Moszoro [39] highlighted that public-private capital structure was the key factor affecting the performance of PPP projects and there should be an optimal capital structure of the public-private shareholdings to achieve maximum efficiency. Therefore, it is necessary to further investigate the impact of diversified PPP project characteristics to identify the project-level drivers of transportation PPP investment affecting the low-carbon economy, which can provide references for the effective design of PPP projects.

3. Data, Variables and Model

3.1. Data Sources

For measuring the level and efficiency of low-carbon economic development, the carbon emissions of 276 cities from 2006 to 2018 are obtained by calculation in this paper. The energy consumption data, energy carbon emission factors, and relevant standards used derive from the IPCC Guidelines for National Greenhouse Gas Inventories 2019 and the China National Standard “General rules for the calculation of the comprehensive energy consumption”. Urban economic data come from the China City Statistical Yearbook, National Economy and Social Development Statistics Bulletin of cities.

In empirically examining the impact of transportation PPP investment on low-carbon economic development, PPP investment data, including railway, road, airport, subway, and waterway, are hand-collected from the China Ministry of Finance Public-Private Partnerships Center and the World Bank Private Participation in Infrastructure Database. To ensure the validity and reliability of the analysis, only executed and successful PPP projects are used.

In the analysis of the impact of PPP project factors, project-level data are all hand-collected from the China Ministry of Finance Public-Private Partnerships Center, publicly available documents pertaining to each project, and the China Government Procurement Platform.

3.2. Research Variables

3.2.1. Explained Variables

A low-carbon economic level and efficiency are core explained variables. The former is measured by carbon productivity, that is GDP produced per unit of carbon dioxide emitted. The latter is the ratio between actual economic output and maximum output under the given factors input in the development of low-carbon economy, which is calculated by the production model. The combination of these two indicators can fully reflect the benefits and efficiency of low-carbon economic development.

The low-carbon economic development level is calculated by Formula (1).

$${\mathrm{CG}}_{\mathrm{it}}{=\mathrm{GDP}}_{\mathrm{it}}/{\mathrm{CO}}_{2\mathrm{it}}$$

(1)

In Formula (1), ${\mathrm{CG}}_{\mathrm{it}}$ is the low-carbon economic development level of the city i in year t. ${\mathrm{GDP}}_{\mathrm{it}}$ is the gross domestic product, which has been converted to a real value based on the data from 2006. ${\mathrm{CO}}_{2\mathrm{it}}$ is the total amount of carbon dioxide emissions. The calculation of total amount of carbon dioxide emissions is based on the general approach of Chinese scholars [40,41], taking the consumption of electricity, coal gas, liquefied petroleum gas, and raw coal in cities, multiplied by its corresponding carbon emission coefficient, and then summed.

Stochastic frontier analysis is adopted to measure the low-carbon economic efficiency of cities. Considering that carbon emissions are a kind of undesired output, regarding them as production input factor as important as capital and labor into the production model conforms to the dual goals of maximizing desired outputs and minimizing undesired outputs in the low-carbon economy [42,43]. Based on the C–D production function [44], and adding carbon emissions as input factor, according to the time-varying model principle proposed by Battese and Coelli [45], the basic measure model was constructed.

$${\mathrm{LnGDP}}_{\mathrm{it}}{=\mathsf{\alpha }}_0{+\mathsf{\alpha }}_1{\mathrm{LnC}}_{\mathrm{it}}{+\mathsf{\alpha }}_2{\mathrm{LnK}}_{\mathrm{it}}{+\mathsf{\alpha }}_3{\mathrm{LnL}}_{\mathrm{it}}{+\mathrm{v}}_{\mathrm{it}}-{\mathrm{u}}_{\mathrm{it}}$$

(2)

In the C–D production function, the output elasticity of each input factor is a fixed constant, and the substitution elasticity obeys the assumption of constant return to scale, which is not in line with economic reality [46]. A trans-log production function overcomes this drawback and more accurately confirms to the economic reality [47]. Therefore, referring to the general form of the trans-log production function under multiple inputs, based on Formula (2), the trans-log production model was constructed.

$$\begin{array}{c}{\mathrm{LnGDP}}_{\mathrm{it}}{=\mathsf{\alpha }}_0{+\mathsf{\alpha }}_1{\mathrm{LnC}}_{\mathrm{it}}{+\mathsf{\alpha }}_2{\mathrm{LnK}}_{\mathrm{it}}{+\mathsf{\alpha }}_3{\mathrm{LnL}}_{\mathrm{it}}{+\mathsf{\alpha }}_4{\mathrm{t}+\mathsf{\alpha }}_5{\mathrm{t}}^2{+\mathsf{\alpha }}_6{\mathrm{LnK}}_{\mathrm{it}}{\mathrm{LnL}}_{\mathrm{it}}{+\mathsf{\alpha }}_7{\mathrm{LnK}}_{\mathrm{it}}{\mathrm{LnC}}_{\mathrm{it}}{+\mathsf{\alpha }}_8{\mathrm{LnL}}_{\mathrm{it}}{\mathrm{LnC}}_{\mathrm{it}}\\ {+\mathsf{\alpha }}_9{\left({\mathrm{LnC}}_{\mathrm{it}}\right)}^2{+\mathsf{\alpha }}_{10}{\left({\mathrm{LnK}}_{\mathrm{it}}\right)}^2{+\mathsf{\alpha }}_{11}{\left({\mathrm{LnL}}_{\mathrm{it}}\right)}^2{+\mathsf{\alpha }}_{12}\left({\mathrm{LnC}}_{\mathrm{it}}\right){\mathrm{t}+\mathsf{\alpha }}_{13}\left({\mathrm{LnK}}_{\mathrm{it}}\right){\mathrm{t}+\mathsf{\alpha }}_{14}\left({\mathrm{LnL}}_{\mathrm{it}}\right){\mathrm{t}+\mathrm{v}}_{\mathrm{it}}-{\mathrm{u}}_{\mathrm{it}}\end{array}$$

(3)

In Formulas (2) and (3), ${\mathrm{GDP}}_{\mathrm{it}}$ is the economic output. ${\mathrm{C}}_{\mathrm{it}}$ represents the carbon emissions. ${\mathrm{K}}_{\mathrm{it}}$ is the capital input, measured by fixed asset investment, which has been converted to a real value based on the data from 2006. ${\mathrm{L}}_{\mathrm{it}}$ is the labor input, measured by the labor force population. ${\mathsf{\alpha }}_1$, ${\mathsf{\alpha }}_2,$ and ${\mathsf{\alpha }}_3$ respectively represent the output elasticity of carbon emissions, capital, and labor. ${\mathrm{v}}_{\mathrm{it}}-{\mathrm{u}}_{\mathrm{it}}$ is a composite error term.

$${\mathrm{u}}_{\mathrm{it}}{=\mathrm{e}}^{-\mathsf{\eta }\left(\mathrm{t}-\mathrm{T}\right)}{\mathrm{u}}_{\mathrm{i}}$$

(4)

In Formula (4), $\mathsf{\eta }$ is a time-varying parameter, for which a value greater than 0 indicates that efficiency tends to improve over time.

Low-carbon economic efficiency is defined as Formula (5).

$${\mathrm{CE}}_{\mathrm{it}}{=\mathrm{e}}^{-{\mathrm{u}}_{\mathrm{it}}}$$

(5)

In Formula (5), ${\mathrm{CE}}_{\mathrm{it}}$ is the low-carbon economic efficiency of the city i in year t. When u = 0 and CE = 1, the individual actual output is equal to the maximum output; this is the state of optimal efficiency. When u > 0 and 0 < CE < 1, this indicates the individual actual output is less than the maximum output and there is a state of efficiency loss.

3.2.2. Explanatory Variables

The core explanatory variable is transportation PPP investment, which is measured by the total amount of transportation PPP investment, including railway, road, airport, subway, and waterways in a city for a year. The characteristic factors of PPP projects are used to further explore the impact on the low-carbon economy at the PPP projects level. Based on data availability, this paper ultimately takes the franchise period, government shareholding proportion, whether social capital is a state-owned enterprise, whether social capital is a listed company, the number of industries involved in the social capital, and investment return mechanism of the project as the explanatory variables.

3.2.3. Mediating Variables

According to the theoretical analysis, there may be a chain relationship between PPP and the low-carbon economy through the upgrading of the industrial structure. This paper uses industrial structure supererogation to measure industrial transformation and upgrading. The upgrading of industrial structure should not only reflect the structural optimization of the contribution of each industry to GDP, but also reflect the improvement of labor productivity of each industry [48]. On this basis, the calculation formula was constructed.

$${\mathrm{IS}}_{\mathrm{it}}={\displaystyle \sum }_{\mathrm{m}=1}^3{\mathrm{Y}}_{\mathrm{it},\mathrm{m}}{\times \mathrm{LP}}_{\mathrm{it},\mathrm{m}},\mathrm{m}=1,2,3$$

(6)

In Formula (6), ${\mathrm{IS}}_{\mathrm{it}}$ is industrial structure supererogation of the city i in year t. ${\mathrm{Y}}_{\mathrm{it},\mathrm{m}}$ represents the proportion of each industry within the GDP. ${\mathrm{LP}}_{\mathrm{it},\mathrm{m}}$ is the labor productivity of each industry; that is, divide each industry’s added value of GDP by the proportion of its employment in total labor force. M 1, 2, and 3 stand for the primary, secondary, and tertiary industries, respectively.

3.2.4. Control Variables

Referring to relevant studies [43,49], this paper takes the economic base, fiscal self-sufficiency rate, degree of government intervention, opening to the outside world, and population size as control variables. The meaning and calculation, and descriptive statistics of all research variables are shown in

Table 1. Meaning and calculation of variables.

Variables	Meaning	Calculation
GDP	Gross domestic product	Take logarithm
C	Carbon dioxide emissions	Take logarithm
K	Fixed assets investment	Take logarithm
L	Labor force population	Take logarithm
CG	Low-carbon economic development level	Divide GDP by carbon emissions
CE	Low-carbon economic efficiency	Use SFA method
PPP	Transportation PPP investment	Take logarithm
FRP	Franchise period in year	Sum of construction and operation periods
GSP	Government shareholding proportion	Divide government share by total share of PPP project companies
SOE	Whether social capital is state-owned	State-owned = 1, otherwise 0
LTE	Whether social capital is listed	Listed = 1, otherwise 0
SIN	Number of industries in the social capital	Sum of industries involved in a PPP project
IEM	Investment return mechanism	Government payment = 1, government subsidy = 2, user payment = 3
IS	Industrial structure supererogation	Use Formula (6)
PG	GDP per capita	Divide GDP by population size
FS	Fiscal self-sufficiency rate	Divide fiscal revenue by expenditure
GI	Government intervention degree	Divide fiscal revenue by GDP
OP	Opening to the outside world	Divide total import and export by GDP
PO	Population size	Take logarithm

and

Table 2. Descriptive statistics of variables.

Variables	Observations	Mean	Sd	Min	Max
GDP	3588	15.702	0.921	13.160	18.686
C	3588	3.031	0.780	0.613	5.441
K	3588	15.831	1.069	12.709	19.045
L	3588	12.756	0.833	10.648	16.105
CG	3588	0.648	0.390	0.087	2.690
CE	3588	0.369	0.148	0.083	1.000
PPP	3588	1.580	4.166	0.000	16.604
FRP	778	19.704	8.631	8	45
GSP	757	0.162	0.183	0.000	1.000
SOE	751	0.872	0.334	0	1
LTE	753	0.206	0.405	0	1
SIN	749	1.559	0.739	1	4
IEM	770	1.566	0.596	1	3
IS	3588	0.302	0.173	0.045	4.026
PG	3588	9.855	0.600	7.926	11.808
FS	3588	0.480	0.231	0.054	1.541
GI	3588	0.071	0.028	0.019	0.238
OP	3588	0.203	0.391	0.000	8.134
PO	3588	5.868	0.692	2.869	8.131

, respectively.

3.3. Model Specification

3.3.1. Model for Examining the Overall Impact of Transportation PPP Investment

This paper first sets up Formula (7) to investigate the overall impact of transportation PPP investment on the low-carbon economy from the dual perspectives of benefits and efficiency. According to the results of the Hausman test, combined with the nature and characteristics of low-carbon economy, the two-factor fixed effects model was constructed.

$${\mathrm{CG}}_{\mathrm{it}}\left({\mathrm{CE}}_{\mathrm{it}}\right){=\mathsf{\alpha }}_0{+\mathsf{\alpha }}_1{\mathrm{PPP}}_{\mathrm{it}}{+\mathsf{\alpha }}_2{\mathrm{PG}}_{\mathrm{it}}{+\mathsf{\alpha }}_3{\mathrm{FS}}_{\mathrm{it}}{+\mathsf{\alpha }}_4{\mathrm{GI}}_{\mathrm{it}}{+\mathsf{\alpha }}_5{\mathrm{OP}}_{\mathrm{it}}{+\mathsf{\alpha }}_6{\mathrm{PO}}_{\mathrm{it}}{+\mathsf{\gamma }}_{\mathrm{i}}{+\mathsf{\delta }}_{\mathrm{t}}$$

(7)

In Formula (7), ${\mathrm{CG}}_{\mathrm{it}}$ and ${\mathrm{CE}}_{\mathrm{it}}$ are the low-carbon economic level and efficiency, respectively; ${\mathrm{PPP}}_{\mathrm{it}}$ is transportation PPP investment (its coefficient measures the net effect of PPP on the low-carbon economy); i represents the city, and t represents the year; and ${\mathsf{\gamma }}_{\mathrm{i}}$, and ${\mathsf{\delta }}_{\mathrm{t}}$ represent the fixed effect of city and year, respectively.

3.3.2. Model for Examining the Spatial Spillover Effect of Transportation PPP Investment

The spatial relationship between variables is not considered in Formula (7), which may lead to potential bias in parameter estimations and not enough in-depth conclusions. The spatial effect is an indispensable analytical perspective which can further extend the relationship between variables to regional interaction, thus providing references for regional integration development. The spatial weight matrix is the basis of spatial study, the scientific selection of which directly affects the accuracy of the findings [50]. The spatial weight matrices can be divided into three basic types: the geographic adjacency matrix, geographic distance matrix, and economic distance matrix. In view of the physical properties of transportation infrastructure interconnection and characteristics of neighborhood cooperation of the PPP mode, this paper selects the geographic adjacency matrix to investigate the spatial correlation between transportation PPP investment and the low-carbon economy, and defines the elements in the matrix as whether two cities are adjacent to each other (adjacent = 1, otherwise 0). Through these calculations, the 276 × 276 dimensional matrix is obtained. On the basis of Formula (7), the spatial Durbin model was constructed.

$${\mathrm{CG}}_{\mathrm{it}}\left({\mathrm{CE}}_{\mathrm{it}}\right)=\mathsf{\rho }{{\displaystyle \sum }}_{\mathrm{j}=1}^{276}{\mathrm{W}}_{\mathrm{ij}}{\times \mathrm{CG}}_{\mathrm{jt}}\left({\mathrm{CE}}_{\mathrm{jt}}\right){+\mathsf{\alpha }}_1{\mathrm{PPP}}_{\mathrm{it}}+{{\displaystyle \sum }}^{}{\mathsf{\alpha }}_{\mathrm{k}}{\mathrm{Control}}_{\mathrm{it}}{+\mathsf{\beta }}_1{{\displaystyle \sum }}_{\mathrm{j}=1}^{276}{\mathrm{W}}_{\mathrm{ij}}{\times \mathrm{PPP}}_{\mathrm{jt}}+{{\displaystyle \sum }}^{}{\mathsf{\beta }}_{\mathrm{k}}{{\displaystyle \sum }}_{\mathrm{j}=1}^{276}{\mathrm{W}}_{\mathrm{ij}}{\times \mathrm{Control}}_{\mathrm{jt}}{+\mathsf{\gamma }}_{\mathrm{i}}{+\mathsf{\delta }}_{\mathrm{t}}$$

(8)

In Formula (8), ρ is the spatial autocorrelation coefficient, showing the spatial dependence between urban low-carbon economies; W represents the 276 × 276 dimensional geographic adjacency matrix; α and β measure the impact of variables on the low-carbon economic development of the local and neighboring cities, respectively; and β₁ measures the spatial spillover effect of transportation PPP investment on the low-carbon economy.

3.3.3. Model for Examining the Mediating Effect of Industrial Structure Supererogation

In order to verify the mediating effect of industrial structure supererogation on transportation PPP investment and low-carbon economic development, the stepwise regression models were constructed.

$${\mathrm{IS}}_{\mathrm{it}}{=\mathrm{b}}_0{+\mathrm{b}}_1{\mathrm{PPP}}_{\mathrm{it}}{+\mathrm{b}}_2{\mathrm{PG}}_{\mathrm{it}}{+\mathrm{b}}_3{\mathrm{FS}}_{\mathrm{it}}{+\mathrm{b}}_4{\mathrm{GI}}_{\mathrm{it}}{+\mathrm{b}}_5{\mathrm{OP}}_{\mathrm{it}}{+\mathrm{b}}_6{\mathrm{PO}}_{\mathrm{it}}{+\mathsf{\gamma }}_{\mathrm{i}}{+\mathsf{\delta }}_{\mathrm{t}}$$

(9)

$${\mathrm{CG}}_{\mathrm{it}}\left({\mathrm{CE}}_{\mathrm{it}}\right){=\mathrm{c}}_0{+\mathrm{c}}_1{\mathrm{PPP}}_{\mathrm{it}}{+\mathrm{c}}_2{\mathrm{IS}}_{\mathrm{it}}{+\mathrm{c}}_3{\mathrm{PG}}_{\mathrm{it}}{+\mathrm{c}}_4{\mathrm{FS}}_{\mathrm{it}}{+\mathrm{c}}_5{\mathrm{GI}}_{\mathrm{it}}{+\mathrm{c}}_6{\mathrm{OP}}_{\mathrm{it}}{+\mathrm{c}}_7{\mathrm{PO}}_{\mathrm{it}}{+\mathsf{\gamma }}_{\mathrm{i}}{+\mathsf{\delta }}_{\mathrm{t}}$$

(10)

In Formulas (9) and (10), ${\mathrm{b}}_1$ measures the impact of transportation PPP investment on industrial structure supererogation. ${\mathrm{c}}_2$ measures the impact of industrial structure supererogation on the low-carbon economy. If these two coefficients are both significant, it can be inferred that the mediating effect holds. If one is not significant, further tests are required.

For further exploring whether there is a spatial relationship of the mediating effect, the spatial stepwise regression models were further constructed based on Formula (8).

$${\mathrm{IS}}_{\mathrm{it}}=\mathsf{\rho }{{\displaystyle \sum }}_{\mathrm{j}=1}^{276}{\mathrm{W}}_{\mathrm{ij}}{\times \mathrm{IS}}_{\mathrm{jt}}{+\mathrm{b}}_1{\mathrm{PPP}}_{\mathrm{it}}+{{\displaystyle \sum }}^{}{\mathrm{b}}_{\mathrm{k}}{\mathrm{Control}}_{\mathrm{it}}{+\mathsf{\epsilon }}_1{\displaystyle {{\displaystyle \sum }}_{\mathrm{j}=1}^{276}}{\mathrm{W}}_{\mathrm{ij}}{\times \mathrm{PPP}}_{\mathrm{jt}}+{{\displaystyle \sum }}^{}{\mathsf{\epsilon }}_{\mathrm{k}}{{\displaystyle \sum }}_{\mathrm{j}=1}^{276}{\mathrm{W}}_{\mathrm{ij}}{\times \mathrm{Control}}_{\mathrm{jt}}{+\mathsf{\gamma }}_{\mathrm{i}}{+\mathsf{\delta }}_{\mathrm{t}}$$

(11)

$$\begin{array}{c}{\mathrm{CG}}_{\mathrm{it}}\left({\mathrm{CE}}_{\mathrm{it}}\right)=\mathsf{\rho }{{\displaystyle \sum }}_{\mathrm{j}=1}^{276}{\mathrm{W}}_{\mathrm{ij}}{\times \mathrm{CG}}_{\mathrm{jt}}\left({\mathrm{CE}}_{\mathrm{jt}}\right){+\mathrm{c}}_1{\mathrm{PPP}}_{\mathrm{it}}{+\mathrm{c}}_2{\mathrm{IS}}_{\mathrm{it}}+{{\displaystyle \sum }}^{}{\mathrm{c}}_{\mathrm{k}}{\mathrm{Control}}_{\mathrm{it}}{+\mathsf{\theta }}_1{{\displaystyle \sum }}_{\mathrm{j}=1}^{276}{\mathrm{W}}_{\mathrm{ij}}{\times \mathrm{PPP}}_{\mathrm{jt}}\\ {+\mathsf{\theta }}_2{{\displaystyle \sum }}_{\mathrm{j}=1}^{276}{\mathrm{W}}_{\mathrm{ij}}{\times \mathrm{IS}}_{\mathrm{jt}}+{{\displaystyle \sum }}^{}{\mathsf{\theta }}_{\mathrm{k}}{{\displaystyle \sum }}_{\mathrm{j}=1}^{276}{\mathrm{W}}_{\mathrm{ij}}{\times \mathrm{Control}}_{\mathrm{jt}}{+\mathsf{\gamma }}_{\mathrm{i}}{+\mathsf{\delta }}_{\mathrm{t}}\end{array}$$

(12)

In Formulas (11) and (12), ${\mathsf{\epsilon }}_1$ measures the spatial spillover effect of transportation PPP investment on industrial structure supererogation. ${\mathsf{\theta }}_2$ measures the spatial spillover effect of industrial structure supererogation on the low-carbon economy. If these two coefficients are both significant, the mediating effect can be further interpreted as the spatial interaction mechanism that transportation PPP investment can affect the upgrading of regional industrial structure, thus affecting regional low-carbon economic development.

3.3.4. Model for Examining the Impact of Characteristics of Transportation PPP Projects

In order to explore the impact of PPP project characteristics on the low-carbon economy, according to the selected project factors, the model was constructed.

$$\mathrm{CG}\left(\mathrm{CE}\right){=\mathsf{\alpha }}_0{+\mathsf{\alpha }}_1{\mathrm{FRP}+\mathsf{\alpha }}_2{\mathrm{GSP}+\mathsf{\alpha }}_3{\mathrm{SOE}+\mathsf{\alpha }}_4{\mathrm{LTE}+\mathsf{\alpha }}_5{\mathrm{SIN}+\mathsf{\alpha }}_6\mathrm{IEM}$$

(13)

In Formula (13), $\mathrm{SOE}$ is a dummy variable, where SOE = 1 indicates social capital includes state-owned enterprises; $\mathrm{LTE}$ is also a dummy variable, where LTE = 1 indicates that the social capital includes listed companies. $\mathrm{IEM}$ is a categorical variable, the return mechanism of PPP projects in China includes government payments, government subsidies, and user payments. This paper defines IEM = 1 for the government-payment mode, IEM = 2 for the government-subsidy mode, and IEM = 3 for the user-payment mode.

A typical feature of PPP projects in China is that state-owned enterprises are the main body of the social capital. In addition, the participation of listed companies in PPP has become a development trend in recent years, which is strongly supported by the Chinese government. In this context, this paper further verifies the impact of state-owned enterprises’ and listed companies’ participation under different PPP project scales. The model to examine the moderating effect was constructed.

$$\mathrm{CG}\left(\mathrm{CE}\right){=\mathsf{\alpha }}_0{+\mathsf{\alpha }}_1{\mathrm{SOE}+\mathsf{\alpha }}_2{\mathrm{PPP}+\mathsf{\alpha }}_3\left(\mathrm{SOE}\times \mathrm{PPP}\right){+\mathsf{\alpha }}_4{\mathrm{FRP}+\mathsf{\alpha }}_5{\mathrm{GSP}+\mathsf{\alpha }}_6{\mathrm{LTE}+\mathsf{\alpha }}_7{\mathrm{SIN}+\mathsf{\alpha }}_8\mathrm{IEM}$$

(14)

$$\mathrm{CG}\left(\mathrm{CE}\right){=\mathsf{\alpha }}_0{+\mathsf{\alpha }}_1{\mathrm{LTE}+\mathsf{\alpha }}_2{\mathrm{PPP}+\mathsf{\alpha }}_3\left(\mathrm{LTE}\times \mathrm{PPP}\right){+\mathsf{\alpha }}_4{\mathrm{FRP}+\mathsf{\alpha }}_5{\mathrm{GSP}+\mathsf{\alpha }}_6{\mathrm{SOE}+\mathsf{\alpha }}_7{\mathrm{SIN}+\mathsf{\alpha }}_8\mathrm{IEM}$$

(15)

In Formulas (14) and (15), $\mathrm{PPP}$ represents the investment scale of a single PPP project; $\mathrm{SOE}\times \mathrm{PPP}$ and $\mathrm{LTE}\times \mathrm{PPP}$ are the interaction terms; and the coefficients of these two variables measure the moderating effect of PPP project investment scale on state-owned enterprise and listed company participation and the low-carbon economy. The project investment scale in the interaction term was mean-centralized in order to overcome the multicollinearity.

4. Empirical Analysis

4.1. Evaluation and Analysis of China’s Low-Carbon Economic Development

Formulas (1)–(5) are used to measure the low-carbon economic development level and efficiency of 276 cities from 2006 to 2018. As the results cover 3588 pieces of data, it is impossible to display all the results in detail. Consequently, this paper focuses on the analysis of the development trend, regional differences, and urban differences. These results are shown in

Table 3. Low-carbon economic development in China from 2006 to 2018.

Year	Level	Efficiency	Correlation Coefficient
2006	0.377	0.351	0.847
2007	0.424	0.354	0.844
2008	0.476	0.357	0.851
2009	0.490	0.360	0.856
2010	0.539	0.363	0.857
2011	0.582	0.366	0.872
2012	0.636	0.369	0.871
2013	0.691	0.372	0.887
2014	0.726	0.375	0.883
2015	0.810	0.378	0.859
2016	0.834	0.381	0.837
2017	0.882	0.384	0.847
2018	0.952	0.387	0.833

Note: The level and efficiency of the low-carbon economy are the average value, same below.

,

Table 4. Low-carbon economic development of different regions in China.

Region	Low-Carbon Economic Level	Low-Carbon Economic Efficiency
East	0.766	0.499
Middle	0.590	0.334
West	0.569	0.295
North	0.517	0.360
South	0.729	0.400

and

Table 5. Low-carbon economic development under classification of cities in China.

Criterion	Classification	Level	Efficiency
Provincial administrative divisions	Municipality directly under central government	1.193	0.601
	Provincial capital	0.854	0.478
	Non-provincial capital	0.617	0.353
Comprehensive strength of cities	First-tier	1.917	0.877
	New first-tier	1.041	0.594
	Second-tier	0.805	0.492
	Third-tier	0.639	0.385
	Fourth-tier	0.625	0.356
	Fifth-tier	0.478	0.252

Note: The classification of cities comes from the China City Statistical Yearbook and China’s authoritative data platform “CBN Data”.

, respectively. See Figure 1 for the scatter plot of the low-carbon economic development level and efficiency.

From 2006 to 2018, China’s low-carbon economy shows a positive development trend which is reflected in the sustaining improvement of both benefits and efficiency. Combined with Table 2, the average low-carbon economic efficiency is only 0.369 over the 13 years, which is far from the optimal efficiency (1.000), indicating that the efficiency loss is still prominent in China’s low-carbon economy and problems such as the low efficiency of resource allocation and unreasonable input structures need to be solved urgently. From the perspective of the coordination of development, the correlation coefficient between the level and efficiency of the low-carbon economy is 0.764, indicating that there is a strong positive correlation between the two indicators; the coefficients of each year show a dynamic tendency of fluctuation, but always maintain a high degree of association. These results manifest the connotation of sustainability that its benefits and efficiency are coordinated and unified in the process of achieving a low-carbon economy in China.

China’s regional low-carbon economic development generally presents a distribution law of “high in the east and low in the west, high in the south and low in the north”, revealing remarkable regional development discrepancy and imbalance, which accords with China’s economic reality. The eastern region is the most economically developed region in China, it relies on its advantageous geographical location, strong innovation ability, abundant financial funds, and national policy support to take the lead to achieve high-quality low-carbon economic development. While western China is restricted by factors such as its environment, infrastructure, and technology, whose development pattern is less advanced. This results in inefficient resource utilization and high pollutant emissions, which impedes its sustainable development of a low-carbon economy. Furthermore, the difference between the north and south is mainly related to the industrial structure of each region. Heavy industry is concentrated in northern China, which inevitably leads to high pollution and carbon emissions, while the south of China is dominated by light industry and the service industry. It can be inferred that industrial structure is an important factor affecting the development of low-carbon economy.

It can be found that there are obvious differences in the development of the low-carbon economy in different types of cities. In the subdivisions of provincial administrative divisions, both the level and efficiency of the low-carbon economy in municipalities directly under central government are higher than provincial capital and higher than non-provincial capital. In the subdivisions of the comprehensive strength of cities, both the level and efficiency of low-carbon economy successively decline from first-tier to fifth-tier cities. The reason lies in that the economic foundation, political status, and comprehensive strength of different cities vary greatly, which results in wide differences in urban development ability and conditions. For instance, first-tier cities have the most favorable position in every aspect and there is no doubt that their development is the fastest.

4.2. Overall Impact of Transportation PPP Investment on the Low-Carbon Economy

Before empirical analysis, it is necessary to conduct the unit root test on variables to judge their stationarity in the panel sequence, which is an important basis to ensure the reliability of results. In this paper, the LLC test and IPS test are used in combination to examine the stationarity of the data, and the results are shown in

Table 6. Unit root test for core variables.

Variables	LLC Test		IPS Test
	Adjusted t	p-Value	Z-t-Tilde-Bar	p-Value
CG	−6.745	0.000	−22.899	0.000
CE	−18.728	0.000	−18.307	0.000
PPP	−13.898	0.000	−18.109	0.000
IS	−7.437	0.000	−15.557	0.000
PG	−6.826	0.000	−17.747	0.000
FS	−20.075	0.000	−10.073	0.000
GI	−11.665	0.000	−1.534	0.063
OP	−21.626	0.000	−15.782	0.000
PO	−7.496	0.000	−3.643	0.000

.

The LLC test and IPS test for core variables both reject the null hypothesis that the panel contains the unit root and consider panel data sequence to be stationary. Therefore, this paper uses raw data for regression. Additionally, Formula (7) is used to examine the overall impact of transportation PPP investment on low-carbon economy. The results are shown in

Table 7. Overall impact of transportation PPP investment on the low-carbon economy.

Variables	CG		CE
PPP	0.005 *** (6.85)	0.004 *** (5.98)	0.001 *** (4.16)	0.001 *** (3.24)
PG		0.209 *** (15.06)		0.001 *** (6.26)
FS		0.286 *** (6.44)		0.002 *** (3.63)
GI		−1.590 *** (−6.83)		−0.009 *** (−2.94)
OP		−0.081 *** (−7.87)		0.001 *** (10.75)
PO		0.553 *** (14.86)		−0.002 *** (−4.20)
Year	Control	Control	Control	Control
City	Control	Control	Control	Control
N	3588	3588	3588	3588
R2	0.681	0.732	0.179	0.741

Note: The values in parentheses are t statistics; *** indicates statistical significance of 99%.

.

Regardless of whether control variables are added, transportation PPP investment is always significantly positively correlated with both the level and efficiency of the low-carbon economy at the statistical level of 1%, showing that transportation PPP investment promotes the low-carbon economic development in China from 2006 to 2018, which is manifested as the dual improvement effects of benefits and efficiency. This has laid a solid foundation for further study on the spatial spillover effect of PPP. The results of the control variables show that the economic base of cities is the core driving factor for low-carbon economic development. Besides, fiscal self-sufficiency, government intervention, opening to the outside world, and population size are also important factors affecting the low-carbon economy.

In order to mitigate the bias caused by endogeneity, referring to existing studies [51,52,53], this paper lags the core explanatory variable PPP by one period to take it as an instrumental variable to perform 2SLS regression. The results are shown in

Table 8. 2SLS regression under the instrumental variable.

Variables	PPP	CG	PPP	CE
	First Stage	Second Stage	First Stage	Second Stage
PPP		0.033 *** (2.79)		0.001 ** (2.11)
IV	0.077 *** (3.76)		0.077 *** (3.76)
PG	2.116 *** (5.11)	0.155 *** (4.70)	2.116 *** (5.11)	0.001 (0.75)
FS	−1.451 (−1.10)	0.326 *** (5.29)	−1.451 (−1.10)	0.003 *** (4.45)
GI	17.844 ** (2.57)	−2.169 *** (−5.76)	17.844 ** (2.57)	−0.017 *** (−3.76)
OP	−0.213 (−0.70)	−0.064 *** (−4.67)	−0.213 (−0.70)	0.001 *** (8.01)
PO	0.227 (0.21)	0.532 *** (10.84)	0.227 (0.21)	−0.002 *** (−2.83)
Year	Control	Control	Control	Control
City	Control	Control	Control	Control
N	3312	3312	3312	3312
R2	0.399	0.531	0.399	0.568
F statistic	117.64		117.64

Note: The values in parentheses are t statistics; ***, ** respectively indicate statistical significance of 99%, 95%. The results of the first-stage regression show that IV is positively correlated with PPP at the 1% significance level; and the F statistic is 117.64, which is greater than the critical value of 10.00 for identifying weak instrumental variables, indicating that the instrumental variable selected is effective. In the second stage, PPP is still significantly positive with CG and CE at the statistical level of 1% and 5%, respectively, which is consistent with the above results. Accordingly, it can be assumed that the basic results of this paper are relatively robust and valid after controlling for endogeneity.

.

4.3. Spatial Spillover Effect of Transportation PPP Investment on the Low-Carbon Economy

Before examining the spatial spillover effect of transportation PPP investment on low-carbon economy, it is necessary to verify the spatial autocorrelation of low-carbon economy with Moran’s Index. The results are shown in

Table 9. Spatial autocorrelation of the low-carbon economy.

Year	CG		CE
	Moran’s I	Z	Moran’s I	Z
2006	0.127 ***	21.09	0.130 ***	21.14
2007	0.123 ***	20.54	0.130 ***	21.16
2008	0.120 ***	19.91	0.130 ***	21.18
2009	0.120 ***	19.87	0.130 ***	21.20
2010	0.121 ***	20.11	0.131 ***	21.21
2011	0.109 ***	17.98	0.131 ***	21.23
2012	0.109 ***	17.99	0.131 ***	21.25
2013	0.113 ***	18.65	0.131 ***	21.26
2014	0.126 ***	20.61	0.131 ***	21.28
2015	0.143 ***	23.25	0.131 ***	21.29
2016	0.161 ***	25.99	0.131 ***	21.31
2017	0.167 ***	26.93	0.131 ***	21.33
2018	0.173 ***	27.89	0.132 ***	21.34

Note: *** indicates statistical significance of 99%.

.

Moran’s I of both CG and CE from 2006 to 2018 is always significantly positive, which shows that there is a strong spatial autocorrelation of the low-carbon economy. Therefore, it is meaningful to extend this study from the traditional perspective to the spatial perspective. Formula (8) is used to explore the spatial relationship between transportation PPP investment and the low-carbon economy; the results of the direct effect, indirect effect, and overall effect are summarized as

Table 10. Spatial spillover effect of transportation PPP investment on the low-carbon economy.

Variables	CG			CE
	Direct	Indirect	Total	Direct	Indirect	Total
PPP	0.003 *** (4.81)	0.007 *** (3.14)	0.010 *** (3.74)	0.001 *** (3.46)	0.001 *** (2.61)	0.001 *** (3.37)
PG	0.140 *** (11.42)	0.102 *** (2.96)	0.242 *** (6.47)	0.001 *** (5.38)	−0.001 (−0.49)	0.001 ** (2.54)
FS	0.278 *** (7.70)	0.102 (0.84)	0.380 *** (2.76)	0.002 *** (3.44)	−0.001 (−0.55)	0.001 (0.97)
GI	−1.895 *** (−10.30)	0.706 (1.08)	−1.189 (−1.58)	−0.007 *** (−2.67)	0.004 (0.72)	−0.003 (−0.40)
OP	−0.062 *** (−7.04)	−0.180 *** (−4.90)	−0.242 *** (−5.75)	0.002 *** (10.28)	0.003 *** (9.95)	0.005 *** (11.98)
PO	0.448 *** (13.67)	0.229 * (1.74)	0.677 *** (4.59)	−0.002 *** (−3.62)	−0.001 (−0.79)	−0.003 ** (−2.05)
ρ	0.613 *** (42.17)			0.320 *** (15.21)
Year	Control			Control
City	Control			Control
N	3588			3588
R2	0.546			0.439

Note: The values in parentheses are t statistics; ***, **, * respectively indicate statistical significance of 99%, 95%, and 90%.

.

The regression coefficients of the direct, indirect, and total effect between PPP and CG, and PPP and CE are always significantly positive, showing that transportation PPP investment has significant and positive spatial spillover effect on the low-carbon economy, which is manifested as the double spatial effects of both benefits and efficiency improvement. Namely, transportation PPP investment can not only promote local low-carbon economic development, but also have a positive impact on neighboring cities. Moreover, by comparing the significance of the coefficients, it can be seen that the coefficients are more significant under the direct effect, implying that urban low-carbon economic development is mainly driven by factors internal to the city itself.

4.4. Mediating Effect of Industrial Structure Supererogation

The premise of testing the mediating effect is that transportation PPP investment has a significant impact on low-carbon economic development, which has been verified in Section 4.2. Therefore, Formulas (9) and (10) are used to examine the overall mediating effect of industrial structure supererogation on transportation PPP investment and the low-carbon economy. The results are shown in

Table 11. Overall mediating effect of industrial structure supererogation.

Variables	IS	CG	CE
PPP	0.004 *** (5.59)	0.003 *** (5.58)	0.001 *** (3.08)
IS		0.069 *** (3.99)	0.001 * (1.84)
PG	0.043 *** (3.08)	0.206 *** (14.86)	0.001 *** (6.17)
FS	0.103 ** (2.29)	0.279 *** (6.29)	0.002 *** (3.57)
GI	−0.334 (−1.42)	−1.567 *** (−6.74)	−0.009 *** (−2.90)
OP	−0.008 (−0.75)	−0.081 *** (−7.84)	0.001 *** (10.77)
PO	0.039 (1.03)	0.550 *** (14.82)	−0.002 *** (−4.23)
Year	Control	Control	Control
City	Control	Control	Control
N	3588	3588	3588
R2	0.229	0.733	0.754

Note: The values in parentheses are t statistics; ***, **, * respectively indicate statistical significance of 99%, 95%, and 90%.

.

There is a significant positive correlation between PPP and IS at the 1% statistical level, showing that transportation PPP investment can promote the upgrading of the industrial structure from the traditional mode to being tertiary-industry dominated. Furthermore, IS is also positively correlated with CG and CE, indicating that industrial structure supererogation can also promote low-carbon economic development. Accordingly, it can be considered that there is a chain relationship between transportation PPP investment and the low-carbon economy through the upgrading of the industrial structure. Formulas (11) and (12) are used to further explore the mediating effect of industrial structure supererogation on transportation PPP investment and the low-carbon economy from the spatial perspective. The results are shown in

Table 12. Spatial mediating effect of industrial structure supererogation.

Variables	IS			CG			CE
	Direct	Indirect	Total	Direct	Indirect	Total	Direct	Indirect	Total
PPP	0.003 *** (4.93)	0.005 *** (3.91)	0.008 *** (5.25)	0.002 *** (4.42)	0.005 *** (2.60)	0.007 *** (3.28)	0.001 *** (3.16)	0.001 ** (2.18)	0.001 *** (3.04)
IS				0.030 ** (2.16)	0.247 *** (4.77)	0.277 *** (4.68)	0.001 (0.69)	0.002 *** (3.63)	0.003 *** (3.50)
CV	Yes			Yes			Yes
ρ	0.207 *** (9.62)			0.605 *** (40.99)			0.315 *** (14.90)
Year	Control			Control			Control
City	Control			Control			Control
N	3588			3588			3588
R2	0.137			0.617			0.489

Note: The values in parentheses are t statistics; ***, ** respectively indicate statistical significance of 99%, 95%; CV represents the set of control variables.

.

The spatial autocorrelation coefficient of IS is significantly positive (0.207), which confirms that the upgrading of the regional industrial structure is an integral, dynamic, and mutually promoting process. In addition to the direct effect, the indirect effect of PPP on IS is still significantly positive, showing that PPP can promote the industrial structure supererogation of the neighboring cities. Moreover, IS also has a significant positive indirect impact on CG and CE. All these reveal a deeper conclusion that the mediating effect has a broader scope of action, which is manifested as a spatial conduction mechanism; that is, transportation PPP investment can promote regional low-carbon economic development by promoting the industrial structure supererogation.

4.5. Heterogeneity Effect of Transportation PPP Investment on the Low-Carbon Economy

Based on the analysis of the overall effect and spatial spillover effect of PPP, this paper ulteriorly discusses the heterogeneous effects of transportation PPP investment in different subsectors on the low-carbon economy; Formulas (7) and (8) are used to examine the heterogeneity in overall effect and spatial spillover effect. The results are shown in

Table 13. Industry heterogeneity in the overall effect of transportation PPP investment.

Subsectors	CG	CE
Road	0.003 *** (4.19)	0.001 *** (3.34)
Railway	0.010 *** (7.59)	−0.001 ** (−2.45)
Airport	−0.007 *** (−2.75)	0.001 (1.05)
Waterway	−0.001 (−0.42)	0.001 * (1.85)
Subway	0.009 *** (6.40)	−0.001 *** (−6.28)

Note: The values in parentheses are t statistics; ***, **, * respectively indicate statistical significance of 99%, 95%, and 90%; the results are after adding all control variables.

and

Table 14. Industry heterogeneity in the spatial spillover effect of transportation PPP investment.

Subsectors	CG			CE
	Direct	Indirect	Total	Direct	Indirect	Total
Road	0.002 *** (2.81)	0.008 *** (3.26)	0.010 *** (3.43)	0.001 *** (3.89)	0.001 (0.39)	0.001 * (1.69)
Railway	0.007 *** (5.83)	0.019 *** (4.75)	0.026 *** (5.49)	−0.001 ** (−2.47)	0.001 ** (1.97)	0.001 (1.06)
Airport	−0.006 ** (−2.33)	−0.021 ** (−2.52)	−0.027 *** (−2.65)	0.001 (1.02)	−0.001 (−0.32)	0.001 (0.12)
Waterway	−0.001 (−0.51)	−0.002 (−0.37)	−0.003 (−0.44)	0.001 (1.44)	0.001 (1.56)	0.001 * (1.66)
Subway	0.009 *** (6.53)	0.005 (0.81)	0.014 ** (2.08)	−0.001 *** (−5.99)	−0.001 (−0.59)	−0.001 ** (−2.44)

Note: The values in parentheses are t statistics; ***, **, * respectively indicate statistical significance of 99%, 95%, and 90%; the results are after adding all control variables.

, respectively.

There is significant heterogeneity in the impacts of transportation PPP investment in different subsectors on the low-carbon economy. In general, road PPP has the strongest promotion effects on both low-carbon economic benefits and efficiency. As the most basic and extensive infrastructure of integrated transportation system, road infrastructure has become the first field in China to introduce social capital, with the largest number of projects and the largest scale of investment. By this same token, PPP has a stronger driving effect on road infrastructure, which effectively improves the quality and quantity of roads’ construction and plays a key role in giving play to the interconnecting effect of road infrastructure and improving the overall efficiency of the regional transportation network. Additionally, railway PPP plays a more important role in driving the integrated development of regional low-carbon economies, which embodies the core value of railway infrastructure as a major economic artery in China. Although railways and subways have the characteristics of low-carbon transportation, their average investment scale is the largest and the public welfare of a subway is stronger, which will cause the asymmetry between input and output, thus weakening their contribution to the efficiency of the low-carbon economy. In addition, airport and waterway construction have higher requirements for geographical conditions, resulting in relatively few PPP projects compared to other fields. Consequently, the scale benefits of PPP investment in both fields on the low-carbon economy have not yet emerged. As an inner-city system, a subway does not have spatial spillover effect, which is consistent with the economic reality.

4.6. Lagged Effect of Transportation PPP Investment on the Low-Carbon Economy

Given the long franchise periods of transportation PPP projects and the sustainability of transportation infrastructure, the explanatory variable is lagged three periods to verify the lagged effect of transportation PPP investment on the low-carbon economy. The results are shown in

Table 15. Lagged effect of transportation PPP investment on the low-carbon economy.

Variables	CG	CE
L1.PPP	0.003 *** (3.61)	0.001 ** (2.50)
L2.PPP	0.002 ** (2.08)	0.001 *** (2.96)
L3.PPP	0.002 (1.01)	0.001 *** (4.00)

Note: The values in parentheses are t statistics; ***, ** respectively indicate statistical significance of 99%, 95%; the results are after adding all control variables.

.

On the whole, the positive impact of transportation PPP investment on China’s low-carbon economy shows relatively strong continuity and stability, highlighting the connotation of economic sustainability of transportation PPP investment. In addition, lagged regression can effectively alleviate the endogeneity between variables, thus guaranteeing the reliability and validity of the research conclusions of this paper.

4.7. Impact of Characteristics of Transportation PPP Projects on the Low-Carbon Economy

Formula (13) is used to explore the impact of transportation PPP projects’ characteristics on low-carbon economic development. The results are shown in

Table 16. Impact of transportation PPP projects characteristics on the low-carbon economy.

Variables	CG	CE
FRP	−0.007 *** (−2.69)	−0.002 *** (−2.89)
GSP	−0.224 ** (−2.02)	−0.057 * (−1.68)
SOE	0.124 ** (2.36)	0.009 (0.58)
LTE	0.123 *** (2.81)	0.048 *** (3.57)
SIN	0.038 (1.55)	0.006 (0.73)
IEM	−0.010 (−0.27)	−0.001 (−0.05)
N	735	735

Note: The values in parentheses are t statistics; ***, **, * respectively indicate statistical significance of 99%, 95%, and 90%.

.

According to the regression results, it is found that franchise period, government shareholding proportion, and nature of the social capital (state-owned enterprise and listed company) are the main project-level factors affecting the low-carbon economy. Among them, FRP and GSP are negatively correlated with the low-carbon economy, while SOE and LTE are positively correlated. Given the public welfare nature of transportation infrastructure, most PPP projects cannot cover the huge costs with revenue during operation and require government payments or subsidies. Additionally, the longer the franchise period, the higher the value of government subsidies tends to be, which will result in fiscal unsustainability and other unsustainable factors. Furthermore, the higher the government’s shareholding in the PPP project company, the stronger the government’s intervention in the project operation and management, which may limit the full play of the advantages of social capital endowment, thus affecting the performance of the project. The participation of state-owned enterprises and listed companies can bring more advanced technology, more scientific operation management experience, and more stable funding sources so as to ensure the efficient implementation of PPP projects.

Next, Formulas (14) and (15) are used to examine the moderating effect of PPP projects’ investment scale on the effects of state-owned enterprise and listed company participation. The results are shown in

Table 17. Moderating effect of PPP projects scale on state-owned enterprises’ participation.

Variables	CG	CE
SOE	0.150 *** (2.86)	0.015 (0.93)
PPP	0.033 ** (1.98)	0.014 *** (2.79)
SOE × PPP	0.077 *** (3.69)	0.020 *** (3.15)
CV	Yes	Yes
N	735	735

Note: The values in parentheses are t statistics; ***, ** respectively indicate statistical significance of 99%, 95%; CV represents the set of control variables.

and

Table 18. Moderating effect of PPP projects scale on listed companies’ participation.

Variables	CG	CE
LTE	0.091 * (1.86)	0.042 *** (2.95)
PPP	0.062 *** (4.13)	0.023 *** (4.70)
LTE × PPP	0.006 (0.21)	−0.007 (−0.76)
CV	Yes	Yes
N	735	735

Note: The values in parentheses are t statistics; ***, * respectively indicate statistical significance of 99%, 90%; CV represents the set of control variables.

.

The regression coefficients of SOE × PPP are significantly positive while those of LTE × PPP are not significant, indicating that the investment scale of PPP projects has a moderating effect on the impact of state-owned enterprises. Namely, the participation of state-owned enterprises can play a more effective role in PPP projects with larger scales. The reason for this is that PPP projects with higher investment scale have higher risk and higher requirements for capital scale and stability. State-owned capital can act as a stabilizer to play a stronger leading role in these PPP projects, which is conducive to ensuring the smooth and efficient construction and operation of transportation infrastructure.

5. Discussion

5.1. The Promotion Effect of Transportation PPP Investment on the Low-Carbon Economy

The findings confirm the low-carbon economic advantages of transportation PPP investment and reveal the deeper implications: (1) the combination of carbon emissions and economic growth should be uniformly considered to analyze the effects of transportation PPP investment, which emphasizes that transportation PPP investment actually improves the productive capacity of the economy driven by carbon emissions rather than absolute carbon reduction. (2) There are various dimensions to measure the low-carbon economy, such as decoupling factors and low-carbon competitiveness. The reasons why this paper does not use these metrics are that they cannot directly reflect the results of low-carbon economy and are also not closely related to transportation PPP investment. The output-oriented indicators more accurately accord with the development goals of coordination of economic benefits and efficiency pursued by developing countries in the process of realizing a low-carbon economy [54]. (3) The spatial spillover effect of transportation PPP investment is closely related to the spatial spillover of the economy itself, reflecting the spatial characteristics of economic factors. Moreover, most transportation PPP projects are jointly invested and constructed among regions, which highlights the economic development mode under interregional cooperation and constitutes the practical motivation of the spatial spillover effect of transportation PPP investment. To sum up, conclusions are in line with the reality of economic and social development in China, as well as the particularity and representativeness of PPP applied in the transportation industry.

5.2. Correctly Understanding the Economic Value of PPP for the Low-Carbon Economy

The positive effects of transportation PPP investment on the low-carbon economy in China are verified based on successfully implemented PPP projects. However, in practice, there are also a large number of failed or suspended transportation PPP projects because of failing to realize the expected value, wrong decisions by government, conflicts of interest, unreasonable estimates of returns, and so on. There is no doubt that these projects will lead to waste of resources and loss of value, which becomes the opposite of their original intention. This is a warning that the application of the PPP mode cannot guarantee the achievement efficiency and the economic value of PPP should be viewed rationally, especially for the transportation industry, where investment and risk are both high. As a result, it should attach great importance to strengthen the top-level design of PPP to prevent unfavorable factors and facilitate efficient cooperation under the conditions of a harmony of interests, reasonable transaction structure, effective government supervision, and risk allocation mechanism. Only in this way can we give full play to the advantages of PPP and realize its expected economic value. Secondly, with the world economy threatened by COVID-19, ensuring the stable investment and supply of infrastructure is a major way to counter economic downturn and stabilize social operation. Against the backdrop of China’s fundamental vision of becoming a country with strong transportation network, improving the quality and efficiency of transportation infrastructure has become more of a priority. In this situation, PPP becomes an inevitable choice, which has made outstanding contributions in boosting investment and economic performance despite COVID-19. Consequently, there is still an extreme necessity to promote high-quality PPP investment to form sustaining economic driving force for long-term sustainable development, especially in the global context of low-carbon economic development.

5.3. Limitations of the Study

This study still has some potential limitations. Firstly, this paper takes China’s reality as the context to explore the impact of transportation PPP investment on the low-carbon economy, but how effective it is in other countries remains unstudied. Even so, conclusions still have certain referential significance for other countries to develop transportation PPP investment, especially for developing countries. Secondly, limited by data availability, the impact of failed projects has not been verified; thus, we cannot highlight a contrast with successful projects. Additionally, due to the paper’s space constraints, further discussions on the heterogeneity of regions and project characteristics of transportation PPP investment in different subsectors have not been carried out deeply. All these are worthy of future research attention. In the next step of the study, this paper will take countries along the Belt and Road as the research object to further discuss the impact of transportation PPP investment on the low-carbon economy of these countries, so as to extend the findings of this study.

6. Conclusions

The results show that: (1) transportation PPP investment markedly promotes the low-carbon economic development of China from 2006 to 2018, which is manifested as the dual improvement effects of benefits and efficiency of the low-carbon economy. (2) Transportation PPP investment can not only promote local low-carbon economic development, but also have a positive spatial spillover effect on neighboring cities. (3) There is a chain relationship between transportation PPP investment and the low-carbon economy through the industrial structure. Namely, transportation PPP investment can promote low-carbon economic development by promoting the upgrading of the industrial structure, which is further manifest as the spatial conduction mechanism. (4) The impacts of transportation PPP investment in different subsectors are significantly heterogeneous. Road PPP has the strongest promotion effects on China’s low-carbon economy. Railway PPP contributes more to the regional low-carbon economy. Subway PPP can improve low-carbon economic benefits rather than efficiency. (5) Transportation PPP investment has a continuous and stable effect on promoting the low-carbon economy in China, which highlights the connotation of economic sustainability in the PPP mode. (6) At the micro-project level, the participation of state-owned enterprises and listed companies has a significant positive impact, and state-owned capital is more effective in PPP projects of relatively large scales. While a longer franchise period and a larger proportion of government shareholding in PPP project companies will reduce the contribution of transportation PPP investment to the low-carbon economy on the whole.

In conclusion, it is absolutely necessary to recognize the indispensable role of transportation PPP investment in low-carbon economic development. In the process of realizing the low-carbon economy, it is critical to make full use of high-quality transportation PPP investment to achieve the coordinated and sustainable development of carbon reduction and economic growth under more effective institutional arrangements. Furthermore, led by the example of the Belt and Road Initiative proposed by China, taking the connection of transportation infrastructure as a bridge and linking it to promote the sound development of a low-carbon economy around the world is possible. In addition, with sustainable regional development as the goal, we must foster a new pattern of coordinated low-carbon development among cities through transportation PPP investment. In the top-level design of PPP projects, different types of social capital, including state-owned enterprises and listed companies, should be encouraged to participate and bring in their diversified advantages to empower the performance of projects.