Front. Environ. Sci. Frontiers in Environmental Science Front. Environ. Sci. 2296-665X Frontiers Media S.A. 769608 10.3389/fenvs.2021.769608 Environmental Science Original Research RETRACTED: Assessing the Role of Environmental Expenditures and Green Transport in Emissions Released by Transport: An Application of ARDL Approach Hussain et al. Green Transport Hussain Zahid 1 Miao Cuifen 2 Zhang Weitu 1 Khan Muhammad Kaleem 3 Xia Zhiqing 1 * 1 School of Finance, Qilu University of Technology (Shandong Academy of Sciences), Jinan, China 2 National Institute of International Strategy, Chinese Academy of Social Sciences, Beijing, China 3 School of Economics and Management, East China Jiao tong University, Nanchang, China

Edited by: Umer Shahzad, Anhui University of Finance and Economics, China

Reviewed by: Diogo Ferraz, Universidade Federal de Ouro Preto, Brazil

Tuhin Sengupta, Indian Institute of Management Ranchi, India

Rida Waheed, Jeddah University, Saudi Arabia

 *Correspondence: Zhiqing Xia, xiazhiqing@hotmail.com

This article was submitted to Environmental Economics and Management, a section of the journal Frontiers in Environmental Science

 12 11 2021 07 08 2025 2021 9 769608 02 09 2021 08 10 2021 Copyright © 2021 Hussain, Miao, Zhang, Khan and Xia. 2021 Hussain, Miao, Zhang, Khan and Xia

This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

This study investigates the effects of transport and environmental factors on transport carbon dioxide emissions (TCO2). It employs cross-sectional autoregressive distributed lags for the estimation in the short and long runs and examines the panel time-series data from 2000 to 2020 in the OECD countries. This method allows heterogeneity in the dependencies and slope parameters across the countries. The results demonstrate that road and railway traffic movements increase the amount of TCO2 in the short and long runs. In addition, transport energy consumption is the driving factor in releasing TCO2 in the long run. Moreover, the joint effect of locomotives and transport energy consumption significantly reduces TCO2 in the short run. By contrast, the findings support the argument that environmental expenditures and green transport mitigate TCO2 in the long run. The findings also show an inverted u-shaped relationship between TCO2 and transport energy consumption. With the empirical findings as a basis, we suggest that the OECD countries should reduce traffic movements and enhance the environmental expenditures so that they may produce green transport vehicles to combat environmental issues.

 traffic transport energy consumption environmental expenditures green transport OECD Introduction

The transport sector has gained great attention owing to concern over environment quality. Environment quality is deteriorating because of carbon dioxide (CO2) emissions, the most critical challenge for countries (Ahmad et al., 2021). The transport sector is a major contributor to CO2 emissions, accounting for approximately 23% (Seum et al., 2020; Churchill et al., 2021; Sohail et al., 2021). Globally, CO2 emissions from the transport sector will increase to nearly 60% by 2050 in the absence of effective mitigation measures (ITF, 2019). IEA (2019) also reports that approximately one-third of global CO2 emissions are from the transport sector. Several economic activities such as wide-range development of transport infrastructure, traffic movement of vehicles, population growth, and economic growth have enhanced demand for transport vehicles, which indicate critical threats to sustainable development.

The transport energy consumption in the OECD countries is projected to reach an average rate of 1.2% from 2012 to 2040. Nevertheless, the patterns of transport energy consumption in the OECD countries are well established with vehicle efficiency improvements (Conti et al., 2016). Energy is the most important factor for economic development, particularly in the transport sector (movement of passengers and freights), and the prime reason for environment deterioration. The association between transport, energy use, environment and economic development has been considered a debatable topic in the current century (Mehmood 2021; Habib et al., 2021; Tan et al., 2021).

Traffic movements are also creating challenges for the transport sector and the environment. The passenger- and freight-intensive movements resulted in trillion trips per annum, which increase the demand for locomotives (vehicles). These movements and locomotives substantially contribute to CO2 emissions (Hadavi et al., 2020). Verlinghieri. (2020) argues that traffic movements indirectly affect CO2 emissions through the usage of road and railway locomotives. Likewise, Heinold. (2020) asserts that road and railway locomotives produce a substantial amount of CO2 emissions through energy consumption.

Considerable literature examines the different scenarios of transport modes with regard to CO2 emissions. We seek to answer the following questions: What is the effect of traffic movements and transport energy consumption on CO2 emissions? Do green transport and environmental expenditures affect CO2 emissions?

This study concentrates on the OECD countries for a number of several reasons. First, the OECD countries are high-income economies that significantly contribute to the global economy, with approximately 42.8% of GDP at purchasing power parity. Second, the OECD countries are responsible for producing CO2 emissions, and their transport energy consumption increases at an average rate of 1.4% per annum (from 104 quadrillion Btu to 155 quadrillion Btu) from 2012 to 2040. Furthermore, 55% of global transport energy consumption is by the OECD countries, whereas non-OECD countries account for around 45% (IEA, 2019). Third, the OECD countries are knowledge-based economies OECD. (2018). Billion trips of passengers and freights are made per year, which lead to traffic movements. Usually, traffic movements for international economic activities are toward the OECD countries. Fourth, the OECD countries are serious in controlling the emissions and improving environment quality. They are spending a larger amount of GDP for mitigating CO2 emissions (Petrović and Lobanov, 2020). Fifth, the OECD countries are shifting their transport resources from carbon transport to green transport (e.g. electric vehicles). Therefore, regardless of the determinants of CO2 emissions, whether and how transport energy consumption and green transport affect the association between CO2 emissions and traffic movements, locomotives and environmental government budget remains a gap in the literature. Acknowledging the influence of traffic movements and green transport on CO2 emissions in OECD countries is imperative to identify policy implications for sustainable transport policies (Rafique et al., 2022; Shahzad et al., 2021; OECD).

This study motivates by making numerous contributions to the literature. First, this study does not only investigate the linear effect of transport energy consumption on transport CO2 emissions (TCO2) but also analyzes the nonlinear effect. Second, this study examines the joint effect of transport energy consumption and locomotives on TCO2. Third, this study provides insights into the relationship between traffic movements and TCO2. Fourth, we explore the effect of environmental research and development expenditures (ERDE) on TCO2 and the joint effect of ERDE and environmental taxes on TCO2. In addition, this study examines whether green transport significantly affects TCO2. Fifth, this study makes methodological contribution by applying the cross-sectional autoregressive distributed lags (CS-ARDL) model to examine the relationship between traffic movements, transport energy consumption, ERDE, and green transport on TCO2. In the presence of cross-dependency, heterogeneity, endogeneity problem, nonstationarity, and misspecification bias, the CS-ARDL is a robust method (Zeqiraj et al., 2020). For robustness check, we use a common correlated effects mean group (CCEMG) approach.

The structure of this article is described as follows. Previous studies are presented in section 2. Section 3 provides the theoretical framework, data source, and methodology. The empirical results and discussions are given in section 4. Last, section 5 summarizes the conclusions and policy implications.

 Literature Review

Several decades take into account the attention on climate change and accelerating the degradation of environmental quality, which are the most critical challenges and big threat for the world. Under such circumstances, the governments need to acknowledge the importance of environmental issues and put into practice to counteract them. Therefore, the reasons behind those human activities are being executed e.g., economic growth, trade, energy use, urbanization, and so on. These activities cause environmental issues pertaining to global warming and climate change. For instance, enormous productivity of greenhouse gases (especially CO2 emissions) is increasing the global temperature, pollution, and degrading the natural resources. Consequently, environment quality is being drastically deteriorated over the time period (Shahzad et al., 2020; Polloni-Silva et al., 2021; Polloni-Silva et al., 2021). Thus, we move to debate on association between transport sector and environment as follows.

Transportation has a significant effect on CO2 emissions around the world. Transport and the environment can be studied in different ways. Generally, the growth of the transport sector has resulted in environment quality costs. Studies show that the environmental effects from the transport sector vary depending on transportation methods and regulations.

Recently, substantial research on the association between environment and transport has gained much attention. Churchill et al. (2021) research the effect of transport infrastructure on CO2 emissions using parametric and nonparametric approaches for a panel of OECD countries. Their findings confirm that a 1% increase in transport infrastructure is associated with a 0.4% increase in CO2 emissions. Furthermore, nonparametric estimation suggests a time-varying relationship between transport infrastructure and CO2 emissions, which is positive throughout World War II and up until now. From a critique perspective, transport infrastructure (e.g., construction of road, railway, and airport) is not the only one responsible for CO2 emissions. Traffic activities also play a significant role in the emissions. Thus, this study includes traffic activities (e.g. movement and locomotives) in analyzing the effect on emissions released by the transport sector.

Subsequently, Ángel et al. (2021) focus on the relationship between road transport and CO2 emissions in 22 European countries. Their results unveil that the transport sector is releasing CO2 emissions that account for approximately 27% of the total emissions. Furthermore, road TCO2 are almost entirely determined by (fossil) fuel consumption. Hence, energy use is the main determinant in CO2 emissions, particularly the transport sector. Another explanation by Pani et al. (2021) also reports that freight transport upsurges the greenhouse emissions in the largest countries. The reason is that truck vehicles for forwarding the shipments stimulate the energy demand. Consequently, CO2 emissions are released and deteriorate the environment quality. Similarly, Cardenete and López-Cabaco. (2021) document that transport of cargo is the most effective factor and that more than 30% of all modes of transport contribute to CO2 emissions in Spain. Likewise, Arvin et al. (2021) analyze the correlation between and energy consumption in Germany. The fuel (e.g. gasoline and diesel) used in the transport vehicles upsurges CO2 emissions through traffic locomotives.

Another evidence documented by Umar et al. (2021) highlights the effects of biomass energy consumption and fossil fuel energy consumption on CO2 emissions in the transport sector in the United States. They find that fossil fuel energy has a positive and significant effect on CO2 emissions released by the transport sector. Biomass energy consumption is negatively associated with CO2 emissions though. Furthermore, they note an inverted U-shaped relationship between energy consumption and CO2 emissions.

By contrast, Hussain et al. (2020) document that climate change potential (CO2 emissions) has a negative association with transport infrastructure. Nonetheless, development infrastructure is also a driving factor of CO2 emissions. Interestingly, extreme climate change potential reduces transport activities through critical infrastructure. Transport emissions are also investigated by Ahmed et al. (2020). Their results confirm that energy consumption in economic growth and the road sector increases emissions.

In support of mitigating CO2 emissions, Sohail et al. (2021) emphasize the association between green transport and environment. They find that green transport is the better strategy to reduce CO2 emissions. Electric vehicles reduce the demand for fossil energy consumption whilst increasing electricity demand. However, electricity shortage issues occur in the market. In this situation, resources are shifted from fossil energy to electric vehicles. Consequently, CO2 emissions by the transport sector tend to decrease.

Another study supports green transport aimed at reducing the CO2 emissions. Oryani et al. (2021) argue that renewable electric vehicles are supported to reduce CO2 emissions per capita. A substantial reduction in CO2 emissions is possible by shifting internal combustible engine vehicles to alternative fuel vehicles.

Moreover, some studies provide evidence for reducing CO2 emissions through environmental tax channels. For instance, Bergantino et al. (2021) analyze the effect of taxes on CO2 emissions to improve environment quality. They argue that taxes on cars decrease CO2 emissions because demand for and supply of cars decline simultaneously in the market. Consequently, an increase in sharing of cars reduces CO2 emissions. Khastar et al. (2020) also point out that an adequate environmental tax level poses a mitigating effect on CO2 emissions in the European Union countries. Furthermore, Reaños (2020) argue that carbon tax has a significant effect on CO2 emissions. To reduce CO2 emissions, carbon taxes need to be imposed (at least 30 Euros per tonne of CO2 emissions) on vehicle owners. Price elasticities suggest that additional carbon taxes may stimulate vehicle owners toward intense energy consumption.

Mariano et al. (2016) evaluated the efficiency analysis on the transport logistics performance. They used a nonparametric approach e.g., slacks-based measure of the data envelopment analysis (DEA) with CO2 emissions (treated as an input) and seven outputs e.g., GDP and six LPI components (treated as an output). Furthermore, window analysis and Malmquist index are also employed to evaluate the efficiency levels over the time. In addition, the DEA technique is based on mathematical foundation and has no specific assumption for analysis. It does not estimate the short-run and long-run relationships between the variables. In contrast, this study focuses on econometric approaches that have specific assumptions for analysis, and examines the short-run and long-run relationships between the variables e.g., transport carbon emissions, traffic, environmental expenditures, and green transport, used in this model.

Extant literature also considers the effect of environmental R&D on CO2 emissions released from the transport sector. Substantial research proves that R&D supports the reduction of CO2 emissions and improvement of environment quality. In this context, Petrović and Lobanov et al. (2020) report that the average effect of R&D is negatively associated with CO2 emissions in the OECD countries. On average, a 0.15% decrease in CO2 emissions is due to a 1% increase in R&D expenditures. Their results confirm that a higher level of R&D expenditures can reduce CO2 emissions, but it does not apply to 40% of countries owing to scarce resources. Wang and Zhang. (2020) also find that a 1% increase in R&D expenditures reduces CO2 emissions by 0.8122% in the BRICS countries.

The research on the association between traffic movements and locomotives must be considered in panoramic aspects. Besides, the joint effect of transport energy consumption and traffic locomotives remains unexplored. The nonlinear effect of transport energy consumption is ignored in the previous literature as well. In addition, the nexus between TCO2 and traffic, transport energy consumption, ERDE, and green transport frequently neglects the potential heterogeneity and cross-sectional dependence. Consequently, a substantial gap exists in the prevailing literature. Therefore, the wide indicators of traffic, environmental expenditures, and green transport should be explored with different methods.

 Methodology Theoretical Framework

After reviewing the literature, we develop the theoretical framework to investigate the relationship between the concerned variables. Figure 1 illuminates that traffic is intensely associated with CO2 emissions. Traffic affects CO2 emissions through movements and locomotives. Specifically, road and railway movements (e.g., passengers and freight) are indirectly associated with CO2 emissions. Therefore, road and railway movements happen through locomotives (which require energy and transport-related resources). Consequently, CO2 emissions are released. In addition, transport energy consumption is investigated in this analysis. Transport energy consumption directly influences CO2 emissions.

Theoretical framework.

The empirical model can be estimated as follows: TCO 2 it = β 0 + β 1 ( ROADMOV it ) + β 2 ( RAILMOV it ) + β 3 ( ROADLOCO it ) + β 4 ( RAILLOCO it ) + β 5   ( ECT it ) + ε it

Eq. 1 indicates that TCO2 is a function of road movement, railway movement, road locomotive, railway locomotive, transport energy consumption, interaction of locomotive and transport energy consumption, and square of transport energy consumption. In addition, cross-sections are denoted by “i” (e.g. 35 OECD countries), whereas “t” represents the period from 2000 to 2020. The term “ β ' s ” indicates the intercept and parameters, whereas the error term is denoted by ‘ ε ’ . TCO2 represents the TCO2 measured in tons per year. ROADMOV indicates road movement that is defined as passenger and freight traveled distance per year. Likewise, RAILMOV represents the railway movement of the passengers and freight traveled distance per year. ROADLOCO indicates that locomotive (i.e. truck per year). RAILLOCO also represents the locomotive related to railway (i.e. freight coaches). ECT shows that transport energy consumption is in terms of tons. Specifically, it is the total energy (i.e. fuel, petrol, and diesel) consumed by the road and railway vehicles and locomotives in terms of tons per year. The quadratic term of transport energy consumption is used to examine whether CO2 emissions reduce once transport energy consumption achieves a threshold level. The joint effect of locomotive and transport energy consumption is denoted by the interaction term (LOCO*ECT).

The previous argument shows that road movement is expected to have a positive effect on TCO2 ( β 1 = ∂ TCO 2 ∂ ROADMOV > 0 ) . Likewise, railway movement plays a crucial role in releasing TCO2. A larger volume of freight (million kilogram) is moved to different locations. Hence, railway movement is predicted to have a positive effect on TCO2 ( β 2 = ∂ TCO 2 ∂ ROAILMOV > 0 ) . The road locomotives (e.g., traction engine and diesel engine) extensively move due to a larger amount of passengers and freights. Thus, it is also predicted to have a positive effect on TCO2 ( β 3 = ∂ TCO 2 ∂ ROADLOCO > 0 ) . The rail locomotives (e.g., steam engine and diesel engine) substantially contribute to CO2 emissions released by the transport sector (Cipek et al., 2021). Therefore, rail locomotive is anticipated to have a negative effect on TCO2 ( β 4 = ∂ TCO 2 ∂ RAILLOCO > 0 ) . Mehmood. (2021) argues that energy consumption in the transport sector is a crucial factor in releasing CO2 emissions, and it is also anticipated to have a positive effect on CO2 emissions ( β 5 = ∂ TCO 2 ∂ ECT > 0 ) . TCO 2 it = β 0 + β 1 ( ROADMOV it ) + β 2 ( RAILMOV it ) + β 3 ( ROADLOCO it ) + β 4 ( RAILLOCO it ) + β 5   ( ECT it ) + β 6 ( LOCO*ECT it ) + ε it

This study also investigates the joint effect of locomotive and transport energy consumption on TCO2. Thus, we add the interaction term in the empirical model (Eq. 1) to examine the effect of locomotive on the association between TCO2 and transport energy consumption. Hence, ( β 6 = ∂ TCO 2 ∂ LOCO ∗ ECT < 0 ) .

Subsequently, we include the quadratic term of transport energy consumption in Eq. 3 to estimate the U-shaped or inverted U-shaped between TCO2 and ECT. TCO 2 it = β 0 + β 1 ( ROADMOV it ) + β 2 ( RAILMOV it ) + β 3 ( ROADLOCO it ) + β 4 ( RAILLOCO it ) + β 5   ( ECT it ) + β 6 ( LOCO*TEC it ) + β 7 ( ECT it 2 ) + ε it

Furthermore, it is supposed to have a positive effect on TCO2 ( β 7 = ∂ TCO 2 ∂ ECT 2 > 0 ) .

Interestingly, we add environmental government budget, ERDE, and green transport factors to analyze their effect on TCO2. TCO 2 it = β 0 + β 1 ( ROADMOV it ) + β 2 ( RAILMOV it ) + β 3 ( R O A D L O C O i t ) + β 4 ( R A I L O C O i t ) + β 5 ( EGB it ) + β 6 ( ERDE it ) + β 7   ( ETAX it ) + β 8 ( ROEV it ) + β 9 ( RAEV it ) + ε it

Eq. 4 shows that TCO2 is also a function of environmental government budget (EGB), ERDE, environmental taxes (ETAX), road electric (ROE), and railway electric (RAE). Environmental government budget is another important factor to protect environment quality. Hussain et al. (2020) argue that the governments allocate the national fund for different environmental projects that enable the government bodies to control the greenhouse emissions released from several sectors, particularly the transport sector and related industries. Thus, environmental government budget is anticipated to have a negative effect on transport energy consumption ( β 5 = ∂ TCO 2 ∂ EGB < 0 ) . Likewise, ERDE is an important factor to control TCO2. Wang et al. (2021) and Chishti et al. (2021) argue that ERDE can reduce TCO2 through advancement in technology. ERDE is expected to have a negative effect on TCO2 ( β 6 = ∂ TCO 2 ∂ ERDE < 0 ) . ETAX is a significant factor as well. Li et al. (2021) argue that environmental taxes discourage the production entities or companies in releasing a certain amount of CO2 emissions. Thus, ETAX is predicted to have a negative effect on TCO2 ( β 7 = ∂ TCO 2 ∂ ETAX < 0 ) . In addition, road electric vehicles have a significant effect on TCO2. Hu et al. (2021) argue that road electric vehicles play a crucial role in reducing TCO2. These vehicles are considered a green transport strategy to protect the environment. Thus, it is anticipated to have a negative effect on TCO2 ( β 8 = ∂ TCO 2 ∂ ROEV < 0 ) . Railway electric vehicles also effect TCO2 as an alternative approach. Kejun et al. (2021) debate that railway electric vehicles impede TCO2 pathways. These vehicles are regarded as an important transport measure to control CO2 emissions released by the transport sector. RAE is expected to have a negative effect on TCO2 ( β 9 = ∂ TCO 2 ∂ RAEV < 0 ) . TCO 2 it = β 0 + β 1 ( ROADMOV it ) + β 2 ( RAILMOV it ) + β 3 ( R O A D L O C O i t ) + β 4 ( R A I L O C O i t ) + β 5 ( EGB it ) + β 6 ( ERDE it ) + β 7   ( ETAX it ) + β 8 ( ROEV it ) + β 9 ( RAEV it ) + β 10 ( ERDE it 2 ) + β 11     ( ROEV it 2 ) + ε it 

We extend the empirical model (Eq. 4) to analyze the nonlinear effect of ERDE and ROE, specifically whether the nexus between TCO2, ERDE and ROE is U-shaped or inverted U-shaped. The coefficients of square of ERDE and ROE are predicted to have a negative effect on TCO2 ( β 10 = ∂ TCO 2 ∂ ERDE 2 < 0 ) , ( β 11 = ∂ TCO 2 ∂ ROE 2 < 0 ) .

We also include the interaction term (EGB*ETAX) in the empirical model (Eq. 6) to analyze the joint effect of EGB and ETAX on TCO2. Environmental government budget is interacted with environmental taxes to estimate the joint effect on TCO2 emissions. More precisely, EGB may mitigate the effect on TCO2 emissions through the environmental taxes. Yuelan et al. (2021) also support that environmental government budget has a significant impact on emissions. The reason is that environmental taxes are also main sources of government budget related to environment (Mirović et al., 2021; Rafique et al., 2022). TCO 2 it = β 0 + β 1 ( ROADMOV it ) + β 2 ( RAILMOV it ) + β 3 ( ROADLOCO it ) + β 4 ( RAILOCO it ) + β 5 ( EGB it ) + β 6 ( ERDE it ) + β 7   ( ETAX it ) + β 8 ( ROEV it ) + β 9 ( RAEV it ) + β 10 ( ERDE it 2 ) + β 11     ( ROEV it 2 ) + β 12 ( EGB*ETAX it ) + ε it 

To analyze the joint effect of EGB and ETAX, we predict that the coefficient of the interaction term has a negative effect on TCO2 ( β 12 = ∂ TCO 2 ∂ EGB ∗ ETAX < 0 ) .

 Data and Source

This study uses balanced panel data set from 2000 to 2020 for 35 OECD countries, namely, Australia, Austria, Belgium, Canada, Chile, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Japan, Korea, Latvia, Lithuania, Luxembourg, Mexico, The Netherlands, New Zealand, Norway, Poland, Portugal, Slovak Republic, Slovenia, Spain, Sweden, Switzerland, Turkey, the United Kingdom, and the United States. We employ 11 variables in the empirical models. Table 1 reveals the operational variables and data sources.

Variables and data sources.

Variable Code Measurement Source
Transport-carbon emission TCO2 Million ton OECD
Road Movement ROADMOV Passenger and freight Kilo meter (million) OECD
Rail Movement RAILMOV IV-07 (1000) OECD
Road Locomotive ROADLOCO Number at (31.12) total traction engine (million) OECD
Rail Locomotive RAILLOCO Number at (31.2) total steam engine, diesel (million) OECD
Transport-energy consumption ECT Percentage of total energy consumption OECD
Road-electric vehicle ROEV Million OECD
Rail-electric vehicle RAEV million OECD
Environmental government budget EGB Share of percentage of total EGB WDI
Environmental R&D expenditures ERDE Percentage of GDP WDI
Environmental Tax ETAX Percentage of GDP WDI

Source: author’s calculation.

 Cross-Sectional Dependence Test

We employ a cross-sectional dependence (CD) test on the variables. The reason is that countries are interconnected through multiple aspects such as economic, cultural, political, and social. As a result, dependency may exist. We follow Pesaran’s (2004) CD and scaled LM test. The equation of CD test is given as CD = 2 T N ( N − 1 )   ( ∑ i = 1 N − 1 ∑ j = i + 1 N ρ ˜ ij ) , where ρ ˜ ij indicates pairwise correlation of the cross-sectional residuals that are obtained from augmented Dickey–Fuller (ADF). Therefore, “T” and “N” are indicators of time and cross-section dimensions, respectively.

 Unit Root Test

This study employs cross-sectionally augmented ADF and cross-sectionally augmented IPS test (Pesaran, 2007) to examine the stationarity characteristics. These also can be known as CADF and CIPS, respectively. Furthermore, arguments by Pesaran. (2007), Moon and Perron. (2004), and Bai and Ng (2004) draw a picture. That is, CIPS approach is a second-generation panel unit root test, and it is an efficient method for CD and heterogeneity. The estimated equation is as follows: Δ CA i , t = ∅ i + ∅ i Z I , t − 1 + ∅ i CSA ¯ t − 1 + ∑ i = 0 ρ ∅ iι Δ CSA ¯ t − 1 + ∑ i = 0 ρ ∅ iι Δ CA i , t − 1 + μ it   where cross-section averages are denoted by C S A ¯ t − 1   and Δ CSA ¯ t − 1 . Therefore, the test statistics of CIPS is given as C I P S ˜ =   1 / N ∑ i = 1 n C D F i  

In Eq. 9, CDF represents the cross-sectional ADF.

 Cointegration Test

To investigate the long-run association between the model parameters, we employ the ECM-based cointegration method (Westerlund, 2007). This method is more appropriate than conventional methods such as Pedroni and Kao. The reason is that the Westerlund cointegration method provides unbiased results in the presence of cross-dependency and heterogeneity. The test statistics can be stated as α i ( L ) Δ γ i t = δ 1 i + δ 2 i t + α i ( γ i t − 1 − β i , x i t − 1 + λ i ( L ) ’ v i t + ε i t ) ,   where δ 1 i = α i ( 1 ) ϕ 2 i − α i ϕ 1 i + α i ϕ 2 i , but δ 2 i = − α i   ϕ 2 i . α i represents the error correction term. The test statistics is described as G t = 1 / N ∑ i = 1 N α i ’ S E ( α i ’ ) G α = 1 / N ∑ i = 1 N T i ’ ( α i ’ ) 1 P t =   α ’ S E ( α ’ ) α ’ = P α / T

The error correction parameters ( α ’ ) in Eq. 10 are computed by replacing the value of P α = T α ’ (Eq. 10.4). Hence, the error correction parameter is specified as ( α ’ ) = P α / T that indicates the ratio of error to be corrected each year in the short run for the disequilibrium case.

 Cross-Sectionally Augmented Autoregressive Distributed Lags

We analyze the short- and long-run relationship among the variables (used in the current empirical models) by employing the CS-ARDL model suggested by Chudik and Pesaran. (2015). The framework of this model includes long-run parameters, short-run ones with error correction and cross-sectional mean. It overcomes the issues such as cross-sectional dependence, heterogeneity, and non-stationarity as robust and produces reliable outcomes (Zeqiraj et al., 2020; Ahmad et al., 2021). The model for CS-ARDL is proposed as follows: ΔTCO 2 i , t = ϑ i + ∑ j = 1 ρ ϑ it ΔTCO 2 i , t − 1 + ∑ j = 1 ρ ϑ ’ ij AV i , t − 1 + ∑ j = 1 ρ ϑ it    Z t − j ¯ + ε it , where ΔTCO 2 i , t is the dependent variable, AV i , t − 1 and   Z t ¯ are the independent variables and the averages for cross-sections, respectively.

 Robustness and Causality Tests

After estimation on CS-ARDL, we check the robustness by using the CCEMG of Pesaran. (2006). This test allows parameters to be heterogeneous in the long run. CS-ARDL is criticized due to imposition of homogeneity restriction in the long run, but economies are diverse with regard to economic and social structures. Subsequently, despite the reliable outcomes obtained by the CS-ARDL and CCEMG, the direction of association between observed variables is not analyzed (which is important for policy implications). We then apply the Dumitrescu and Hurlin (D&H, 2012) method to investigate the casual relationship among the variables (used in the current models). This method provides two statistics that are test average ( W ¯ ) and standard normal distribution ( Z ¯ ). The model can be described as Z i , t =   α i + ∑ j = 1 ρ β i j Z i , t − 1 + ∑ j = 1 ρ γ i j T i , t − j   , where β j (j) and j indicate the autoregressive parameters and lag length, respectively.

 Empirical Results and Discussions Descriptive Statistics

Table 2 reveals the descriptive statistics. The mean values and standard deviations of road movement (ROAD-MOV), rail movement (RAIL-MOV), rail locomotive (RAIL-LOCO), road locomotive (ROAD-LOCO), and transport energy consumption (ECT) are very high, indicating high heterogeneity in the variables across the OECD countries. Moreover, the standard deviations of EGB and TCO2 are higher (11.39 and 72.99, respectively), implying that observations follow a skewed distribution across the sample countries. However, the mean value and standard deviation of environmental tax (ETAX) are smaller, indicating that observations vary within a narrow range over the period.

Descriptive statistics.

Variable Obs Mean Std.Dev Min Max
TCO2 735 92.65 72.998 11.2 861
ROAD-MOV 735 84581.4 548000 1 5210000
RAIL—MOV 735 59434.79 182000 1 1140000
RAIL-LOCO 735 59434.79 182000 1 1140000
ROAD-LOCO 735 666000 2950000 1 2.26e+07
ECT 735 253.327 1759.159 1 27615
EGB 735 25.695 11.399 1 58.94
ERDE 735 2.49 2.094 0.05 17.66
ETAX 735 0.104 0.119 0.01 1
ROAD-ELECT 735 20.32 0.123 0.60 652143
RAIL-ELECT 735 10.23 0.432 0.30 32154

Source: author’s calculations.

We investigate the cross-dependency on the observed variables. To analyze the models, detecting the presence of CD is important. Pesaran. (2004) explains that ordinary econometric methods often cannot overcome the bias in the panels because of presence of CD. Table 3 reports the estimate of Pesaran’s CD and scaled LM test. The results reveal that CD is supported by the value of absolute mean (ranging from 0.696 to 107.619). Therefore, the outcomes of Pesaran’s CD and Pesaran scaled LM test are highly significant for the entire observed variable, indicating that variables have CD. It is worth noticing because with globalization, emerging economies are interconnected. Consequently, the outcomes of the CD test are significantly projected in the model. Some possible changes in the observed variables of emerging economies may affect those of the other economies.

Cross-Sectional Dependence.

Pesaran CD Pesaran Scaled LM
Variable CD-test abs (corr) CD-test
TCO2 101.153*** 0.90 132.333***
ROAD-MOV 22.129*** 0.20 231.433***
RAIL—MOV 44.54*** 0.40 154.298***
RAIL-LOCO 28.464*** 0.26 120.938***
ROAD-LOCO 17.447*** 0.16 121.433***
ECT 107.619 *** 0.96 201.432***
EGB 9.149*** 0.30 103.322***
ERDE 4.186*** 0.08 98.322***
ETAX 54.142*** 0.55 123.543***
ECT2 100.157*** 0.90 190.432***
EGB2 5.315*** 0.28 134.329***
ERDE2 3.953*** 0.08 102.433***

Source: author’s calculations.

Note: Table 2 reveals the estimate of cross-dependency (CD) test of Pesaran.

CD and Pesaran Scaled LM of observed variables of 35 OECD countries.

To investigate the order of integration, we employ a second-generation panel unit root test (CIPS and CADF) by Pesaran. (2007), which shows that cross-sectionally unbiased is a primary feature of CIPS. Investigation on the order of integration is a unique factor in the estimation technique. The outcomes of the CIPS and CADF are summarized in Table 4. The findings exhibit that all variables are stationary at 1 and follow a mixed order of integration. The presence of CD and mixed order of integration require the usage of CS-ARDL framework. Subsequently, we employ a Westerlund cointegration approach to investigate the long-run relationship in the models. The outcomes are given in Table 5, which shows that a long-run relationship exists in the models. Furthermore, error correction (EC) can be calculated by P α value in the models. Hence, the parameter of EC is ( α ' = P α T ) = − 3.966 / 18 = − 0.220 for model 1, − 5.893 / 18 = − 0.327 for model 2, and − 3.854 / 18 = − 0.214 for model 3. The errors around 25.37% between TCO2 and its determinants are corrected each year, so disequilibrium in the short run becomes stable in the long run.

Panel Unit Root.

Cross-sectionally Augmented IPS (CIPS) Cross-sectionally Augmented Dicky-Fuller (CADF)
Variable Level First-difference Level Firs-difference
TCO2 −2.498 −1.232*** −4.479*** −11.086***
ROAD-MOV −0.871 −0.432* 10.961 9.678***
RAIL—MOV −1.698 −1.327*** 2.823 −0.799***
RAIL-LOCO −0.190 *** − 12.269 6.966***
ROAD-LOCO −0.252*** − 13.296 12.567***
ECT −2.617 −1.432*** −1.404 −9.930***
EGB −2.614* −1.232*** −1.576 * −9.157***
ERDE −0.530 * −0.432* 14.714 7.925**
ETAX −1.754 −1.422* 3.558 −4.340***
ECT2 −2.610 0.323* −1.609* −10.079***
EGB2 −2.869*** − −2.222 *** −9.548***
ERDE2 −0.516* −0.431* 14.714 7.925**

Source: author’s calculations.

Note: CIPS, CADF.

Cointegration Test.

Variable Model 1 Model 2 Model 3
Gt −2.557*** −1.540 −1.635***
(−5.133) (1.566) (0.938)
Ga −2.636*** −3.354 −3.983*
(4.897) (4.117) (3.433)
Pt −17.093*** −16.099*** −9.259***
(−8.578) (−7.578) (−0.699)
Pa −3.966*** −5.893*** −3.854***
(0.352) (−2.216) (0.501)

Source: author’s calculations.

Note: Westerlund cointegration test.

 Role of Traffic and Transport Energy Consumption

After evaluation of cointegration, we employ a CS-ARDL method to gauge the dynamic effect of traffic and transport energy consumption in the short and long runs. Table 6 displays the estimation using CS-ARDL. The magnitude of each coefficient indicates a significant relationship between explanatory variables and TCO2 in the short and long runs. A 29.5% increase in TCO2 is due to a 1% increase in road movement in the short run, whereas a 4.3% augmentation in TCO2 is affected by a similar rise in rail movement in the short run. Conversely, road and rail movement variables have relatively less influence on TCO2 in the long run. Payus et al. (2019), Chen et al. (2018), and Grote et al. (2018) argue that movement of passengers upsurges the use of road vehicles. Consequently, CO2 emissions released in the environment are due to traffic activities.

The role of traffic and transport energy consumption.

Variable Short-run Long-run
Model 1 Model 2 Model 3 Model 1 Model 2 Model 3
ROAD-MOV 0.183*** 0.295*** 0.254*** 0.091*** 0.181*** 0.127***
(0.232) (0.028) (0.034) (0.023) (0.24) (0.032)
RAIL-MOV 0.110*** 0.043*** 0.098*** 0.543** 0.087*** 0.432**
(0.034) (0.057) (0.032) (0.432) (1.216) (0.321)
ECT 0.403*** 0.314*** 0.323*** 0.220*** 12.6*** 0.201**
(0.332) (0.847) (0.023) (0.329) (0.498) (0.232)
ROAD-LOCO 0.782*** 0.890*** 0.891*** 0.792** 9.576*** 0.702***
(0.143) (0.096) (0.213) (0.321) (9.221) (0.485)
RAIL-LOCO 0.348*** 0.085*** 0.232*** 0.038*** 1.431*** 0.243***
(0.432) (0.078) (0.043) (0.043) (1.048) (0.093)
ECT*LOCOMOTIVE — 0.015* 0.002* — 0.0613* 0.006*
— (0.470) (0.123) — (1.123) (0.083)
ECT2 — — −0.432*** — — −0.543***
— — (0.432) — — (0.093)
SR Error Correction −0.325*** −0.234*** −0.308*** — — —
(0.065) (0.098) (0.088) — — —
Observations 630 630 630 630 630 630
R-squared 0.499 0.639 0.604 0.575 0.475 0.363
Number of groups 35 35 35 35 35 35

Note: *, **, and *** denote 10, 5 and 1% significance level. The standard errors are in parentheses.

Besides, the positive and significant coefficient of ECT indicates an increase in TCO2. Numerous studies (e.g., Adams et al., 2020; Figliozzi 2020; Peng et al., 2020) find that the use of energy is the most influencing factor of TCO2, and a positive relationship exists between energy consumption and CO2 emissions in the transport sector. A higher volume of economic activities particularly in the transport sector enhance energy consumption, which then increases CO2 emissions.

The effect of road and rail locomotive is positive and significant with respect to TCO2, which implies that extensive use of locomotive is causing TCO2. Entities buy more fossil fuel vehicles (e.g., truck, bus, car, and motorcycle), so TCO2 increase. Several other studies (e.g., Rietmann et al., 2020; Arvin et al., 2021; Li et al., 2021; Xu et al., 2020; Yan et al., 2021) endorse this finding, arguing that locomotives are the main driving factors of CO2 emissions in the transport sector. The demand for trip/traveling within a specific time stimulates the use of fuel vehicles by individuals/consumers. Therefore, road and rail locomotive augments TCO2. The coefficient of ECT square indicates that a 1% increase in ECT2 decreases TCO2 by 43.2 and 54.3% in the short and long runs, respectively. An inverted U-shaped relationship occurs, indicating that a particular amount of energy consumption in the transport sector can reduce TCO2 by adopting environmental strategies (e.g. efficient energy use, electric vehicles, and sustainable locomotive production and consumption patterns). Furthermore, the outcome reflects that OECD countries’ environmental strategies are in the right direction as their economies are progressively adopting alternative uses of transport vehicles to protect their environment (Lin, (2020); He et al., 2021; Rietmann et al., 2020; Lin, (2020); Sharma and Chandel. 2020; Harvey 2020). Subsequently, the coefficient of interaction term (ECT*LOCOM) indicates that a joint effect of transport energy consumption and locomotives increases TCO2, implying that locomotives increase TCO2 through ECT.

 Role of Environmental Expenditures and Green Transport

After assessing the effect of traffic and transport energy consumption on TCO2, we can move forward to investigate the role of environmental expenditure and green transport to resolve the issue of CO2 emissions released by the transport sector. The coefficients of environmental government budget (EGB) are negative and significant in Table 7, which indicate that a 1% increase in EGB reduces TCO2 by approximately 146% (model 1), 13.2% (model 2), and 2.1% (model) in the short run. On average, 28.6, 1.7, and 22.1% decreases in TCO2 in all the cases are due to EGB in the long run.

The role of environmental expenditures and green transport.

Variable Short-run Long-run
Model 1 Model 2 Model 3 Model 1 Model 2 Model 3
ROAD-MOV 0.021*** 0.021** −0.048 0.032** 0.048* 0.043***
(0.032) (0.032) (0.094) (0.038) (0.032) (0.052)
RAIL-MOV 0.023*** 0.039*** 0.321* 0.021*** 0.003** 0.031**
(0.032) (0.003) (0.054) (0.043) (0.085) (0.076)
ROAD-LOCO −0.383 0.038*** −0.003** 0.048** −0.039 0.010**
(0.438) (0.083) (0.084) (0.043) (0.998) (0.076)
RAIL-LOCO 0.002*** 0.004* −0.033 0.323*** 0.021* 0.001**
(0.053) (0.021) (0.098) (0.048) (0.443) (0.028)
EGB −0.021*** −0.132*** −1.461*** −0.221*** −0.017*** −0.286***
(0.212) (0.021) (0.691) (0.021) (0.012) (0.211)
ERDE −0.217*** −0.321*** −0.031*** −0.028*** −0.018*** −0.581***
(0.021) (0.032) (0.031) (0.032) (0.028) (3.178)
ETAX −0.021** −0.028** −0.55** −0.021*** −0.038*** −0.468***
(0.081) (0.212) (0.11) (0.028) (0.321) (0.791)
RAIL-ELECTRIC −0.432*** −0.124*** −0.103*** −0.765*** −0.543*** −0.155***
(0.023) (0.082) (0.089) (0.033) (0.032) (0.005)
ROAD-ELECTRIC −0.472*** −0.328*** −0.276*** −0.573*** −0.382*** −0.335***
(0.032) (0.024) (0.231) (0.076) (0.063) (0.155)
ERDE2 — −0.128** −0.323** — −0.053*** −0.281***
— (0.022) (0.082) — (0.032) (0.321)
ROAD-ELECT-SQU — −0.543*** −0.707*** — −0.654*** −0.179***
— (0.032) (0.887) — (0.203) (0.318)
EGB*ETAX — — −0.007*** — — −0.004***
— — (0.841) — — (0.045)
SR Error Correction −0.398*** −0.403*** −0.432*** — — —
(0.083) (0.032) (0.093) — — —
Observations 630 630 630 630 630 630
R-squared 0.513 0.404 0.287 0.482 0.288 0.295
Number of groups 35 35 35 35 35 35

Note: *, **, and *** denote 10, 5 and 1% significance level. The standard errors are in parentheses.

The results reveal that the current environmental policies of the OECD countries are in the right direction for protecting the environment, particularly regarding the release of TCO2. Some related studies (e.g. Fan et al., 2020; Yang et al., 2020) argue that the government budget related to the environment has a remarkable effect on CO2 emissions. It is considered a driving force to tackle environmental issues. Lack of funding can impede the identification of environmental issues though. Thus, the government budget supports the relevant departments or entities to tackle the problems. However, OECD countries can also constrain the governmental budget related to the environment.

ERDE is another important factor to mitigate CO2 emissions. The coefficient of ERDE is negative and significant. On average, a 1% upsurge in ERDE reduces TCO2 by approximately 3.1% (model 1), 32.1% (model 2), and 21.7% (model 3) in the short run. In the long run, a 58.1% decrease in TCO2 is due to ERDE. These outcomes partially coincide with related studies by Wang et al. (2021) and Chishti et al. (2021) who explain the remarkable effect of ERDE on TCO2 because the former drives invention of new technology for tackling the environmental issues.

Additionally, the coefficient of ERDE square explains that a 1% increase in ERDE square decreases TCO2 by approximately 32.3% (model 1), 12.8% (model 2), and 43.2% (model 3) in the short run. The negative relationship between ERDE square and TCO2 is also apparent in the long run in OECD countries. This indicates that TCO2 can be reduced with technological innovation and environmental economic policies after reaching a specific level of R&D. It also shows that OECD countries are giving greater attention on environment by adopting the alternative approaches, such as green transportation through green financing (Kong et al., 2021; Song et al., 2021).

Subsequently, our investigation focuses on the effect of environmental taxes on TCO2. The coefficients of ETAX are negative and significant in all cases, which shows that 55% (model 1), 2.8% (model 2), and 2.1% (model 3) decrease in TCO2 is due to environmental taxes in the short run. A slight change in the coefficients of ETAX is noted, and the negative correlation among the variables is validated in the long run. This finding is also consistent with those of other studies, for instance, Hao et al. (2021), Ma et al. (2021), and Zhai et al. (2021) who document that environmental taxes have a remarkable effect on CO2 emissions. The reason is that taxes are sources of government bodies which are used for mega projects, including sustainable development, agriculture and industrial growth, and infrastructure (Arvin et al., 2021).

Now our concentration moves to the effect of green transport. Electric vehicles are used as a proxy to analyze the effect on TCO2. The coefficients of railway electric vehicles are negative and significant for TCO2 in all the cases. A 1% increase in railway electric vehicles decreases TCO2 by approximately 43.2 and 76.5% in the short and long runs, respectively (see Table 6). On the contrary, the coefficients of road vehicles also validate the negative relationship with TCO2. On average, in the short run, 27.6% decrease in TCO2 for model 1, 32.8% for model 2, and 47.2% for model 3 are due to road vehicles. Furthermore, a 1% increase in road vehicles decreases TCO2 by approximately 33.3% (model 1), 38.2% (model 2), and 57.3% (model 3) in the long run (see Table 6).

An inverted U-shaped relationship between TCO2 and road vehicles shows that after reaching a specific level of road vehicles, TCO2 can be reduced with better transport and environmental policies. OECD countries are progressively shifting their resources from diesel to electric vehicles. Our findings are consistent with those of related studies by Xu et al. (2021) and Zhang and Hanaoka (2021) who conclude that electric vehicles have a remarkable effect on TCO2 as an alternative transport strategy.

The interaction term of EGB and ETAX is negative and significant. The joint effect of EGB and ETAX reduces TCO2, implying that governments simultaneously allocate the budget and impose taxes related to environment to reduce CO2 emissions.

 Endogeneity and Robustness Check

Endogeneity is the concept of econometric that explains about the correlation between explanatory variables and error terms. Distinguish between endogenous and exogenous variables created in simultaneous equations models where a separate variable is determined by the model that is predetermined. Usually, endogeneity problem arises due to correlation between the explanatory variables and errors term from unobserved or omitted variables is confound both independent variable and dependent variable.

Thus, we attempt endogeneity and robustness of the modeling techniques to present the potential reverse causality problem. We use the best strategy, namely, CCEMG, to estimate the correlation of unobserved variables with explanatory variables and error term. This model can correlate with unobserved variables, such as investment in transport, total energy consumption, freight and passenger volume, and CO2 emission release from other sectors. Table 8 reports the estimate, which indicates that the outcomes are significant. The results portray that the outcomes produced by CS-ARDL are valid, indicating consistent findings. In addition, Table 8 (model 1) shows that traffic factors (road movements, rail movements, road locomotives, rail locomotives, and transport energy consumption) are statistically significant at 1% level. It indicates that the relationship is consistent with the main approach. On the other hand, model 2 also reveals that there is no drawback in findings with the CS-ARDL estimation. Likewise, the coefficients of green transport in model 3 also indicate that there is no robust effect in the model. Generally, unobserved or omitted variables are not correlated with explanatory and dependent variable (Hussain et al., 2020; Mehmood et al., 2021; Isik et al., 2021).

Robustness Check.

Model 1 Model 2 Model 3
ROAD-MOV 0.0295*** 0.040*** 3.245**
(0.002) (0.009) (0.026)
RAIL-MOV 0.0438** 0.012*** 0.009***
(0.057) (0.002) (0.002)
ROAD-LOCO 0.085*** 0.096*** −0.112
(0.000) (0.000) (0.091)
RAIL-LOCO 0.089*** 2.210 3.041***
(0.009) (0.505) (0.000)
ECT 0.0382** 0.124*** 0.132***
(0.012) (0.132) (0.023)
EGB — 0.425*** —
— (0.127) —
ERDE — −2.1705* —
— (0.470) —
ETAX — 0.314*** —
— (0.847) —
RAIL-ELECTRIC — — 0.546***
— — (0.127)
ROAD-ELECTRIC — — 0.031***
— — (0.082)
ROAD-ELECT-SQU — — −0.001***
— — (0.142)
EGB*ETAX — -0.002** —
— (0.032) —
Observations 630 630 630
R-squared 0.639 0.316 0.355
Number of groups 35 35 35

Note: *, **, and *** denote 10, 5 and 1% significance level. The standard errors are in parentheses.

We then carry out the D&H panel causality test to estimate the causal association (Dumitrescu and Hurlin, 2012). The outcomes are given in Table 9. Road movement and TCO2 and environmental taxes and TCO2 have bidirectional causality. Therefore, any policy shock in road movement and environmental taxes can substantially affect TCO2 and vice versa. For instance, an increase in CO2 emissions requires higher amount of environmental taxes (that are source of environmental revenue) (Tao et al., 2021). Conversely, one-way causality to TCO2 cannot be reversed among variables such as transport energy consumption, environmental government budget, and road locomotive. Tao and Wu. (2021) also support that any policy change in transport energy consumption cannot affect TCO2 emissions indicating that the transport energy consumption independently does work in its own mechanism without being affected by external factors.

Panel Causality Test Results.

Null Hypothesis W-statistics Zbar-statistics Prob Conclusion
TCO2 ↮ ROADMOV 1.323*** 0.343 0.000
ROADMOV ↮ TCO2 1.237*** 0.432 0.000 ROADMOV → TCO2
TCO2 ↮ RAILMOV 1.543 0.733 0.234
RAILMOV ↮ TCO2 1.454*** 0.873 0.000 RAILMOV → TCO2
TCO2 ↮ ECT 2.1230 0.3638 0.7160
ECT ↮ TCO2 1.4327*** -1.6781 0.0033 ECT → TCO2
TCO2 ↮ ROLOC 1.554 0.432 0.432
ROLOC ↮ TCO2 1.654*** 0.232 0.000 ROLOC → TCO2
TCO2 ↮ EGB 2.006 0.0177 0.9859
EGB ↮ TCO2 3.3591*** 4.0204 0.0001 EGB → TCO2
TCO2 ↮ ERDE 1.343 0.393 0.0898
ERDE ↮ TCO2 1.543*** 0.493 0.000
TCO2 ↮ ETAX 3.4939*** 4.4189 0.0000 ETAX ↔ TCO2
ETAX ↮ TCO2 6.9714*** 14.7057 0.0000

Source: author’s calculations.

Note: “ → ”indicates one-way causality, while “ ↔ ” two-way causality between the variables.

 Conclusion

We investigate the role of traffic and transport energy consumption in the release of TCO2 and examine the effect of environmental expenditures and green transport of OECD countries from 2000 to 2020. We employ the second-generation empirical tools by Pesaran. (2004) and Pesaran and Yamagata (2008) to check the CD and heterogeneity, respectively. We also employ the unit root tests of CIPS and CADF by Pesaran. (2007). To examine the long-run equilibrium association among the variables, we adopt the Westerlund. (2007) cointegration technique.

We find a CD issue presence in the dataset, and the model is suffering from slope heterogeneity. Consequently, the long-run relationship persists between the variables suggested by the cointegration technique. The findings from the CS-ARDL test show that traffic movement (i.e. road and railway) escalates TCO2. Extensive use of road and railway locomotives augments CO2 emissions. The findings further reveal that transport energy consumption has a remarkable effect on CO2 emissions, which means that energy consumption in the transport sector is a driving factor of CO2 emissions. The joint effect of traffic locomotives (road and railway) and transport energy consumption substantially effects TCO2 as well.

Additionally, environmental government budget greatly diminishes TCO2 through the R&D channel, which stimulates the innovations related to environmental mitigation. R&D expenditures generate innovations or new technology. On the contrary, road and railway vehicles (green transport) substantially degrade TCO2. More precisely, electric vehicles are a better alternative strategy to mitigate TCO2 in OECD countries.

Lastly, the causality results (D&H) portray that any policy that targets road movement and environmental taxes significantly changes TCO2 and vice versa. Any policy related to transport energy consumption, road locomotives, and environmental government budget remarkably change TCO2. On the contrary, any supporting policy to TCO2 does not affect road locomotives and environmental government budget.

 Policy Implications

After concluding the findings, this study recommends the following policy implications for transport experts/economists, urban planners, and transport modelers. First, the results reveal that policymakers in OECD countries should take radical steps to moderate the traffic movement, particularly road movements, to curb the deterioration of the environment. The movement of vehicles on the road must be controlled by transport institutes. The movement strategy can reduce CO2 emissions approximately 18.3 and 11% for road and rail movement respectively (ref. Table 6). Second, government bodies or transport experts/policymakers should reduce 78.3 and 19.3% road and rail locomotives to improve better environment. Third, the governments of the OECD countries should increase their budget by 12.1% on average related to the environment to control CO2 emissions (ref. Table 7). A better expenditure strategy related to the environment can be devised to preserve nonrenewable resources. Fourth, green transport (railway and road electric vehicles) should be produced around 59.85 and 52.25% by the OECD countries respectively (ref. Table7). Fifth, the policy related to environmental taxes and R&D expenditures must be planned for the long run to achieve desirable consequences. Sixth, urban planners must design the transport infrastructure at city level, where traffic movements could be to avoid the releasing of CO2 emissions. Seventh, transport modelers can formulate a policy by considering the demand for and supply of products of green transport e.g., electric vehicles. Transport experts (producers) should reduce the production of rail and road locomotives, as these consume a greater amount of energy during movements. Furthermore, they should control traffic movements within the urban areas that cause the transport carbon emissions.

The limitations of this study are as follows: This study is limited to OECD countries. Only road and railway traffic movements, energy use, green transport, and environmental expenditures are considered to estimate the effect on TCO2. More specifically, transport sector has other factors such as transport household expenditure, transport infrastructure stock, transport logistics performance. It can be reasoned that these factors have been entrenched in/by traffic that directly or indirectly cause environmental degradation. In addition, the current study focuses on the OECD countries and it would be appropriate to enlarge the investigation to other countries e.g., BRICS, OBOR, or emerging economies and non-OECD as well. This research can be extended in the future by including environmental technology, agglomeration of population, and institutional quality. In particular, investigating the effects of green financing, green transportation, and green technology including emerging economies can provide policy makers/practitioners with more specific insights.

 Data Availability Statement

Publicly available datasets were analyzed in this study. This data can be found here: https://stats.oecd.org.

 Author Contributions

ZH conceived, carried out the literature review, and drafted the manuscript. MK, XZ, CM and WZ conceived, designed, and coordinated the study, and contributed to and finalized the draft. All authors read and approved the final manuscript.

 Conflict of Interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

 Publisher’s Note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

 References Adams S. Boateng E. Acheampong A. O. (2020). Transport Energy Consumption and Environmental Quality: Does Urbanization Matter. Sci. Total Environ. 744, 140617. 10.1016/j.scitotenv.2020.140617 Ahmad M. Ahmed Z. Majeed A. Huang B. (2021). An Environmental Impact Assessment of Economic Complexity and Energy Consumption: Does Institutional Quality Make a Difference. Environ. Impact Assess. Rev. 89, 106603. 10.1016/j.eiar.2021.106603 Ángel S. M. Marrero G. A. González R. M. Rodríguez-López J. (2021). Convergence in Road Transport CO2 Emissions in Europe. Energ. Econ. 99. 10.1016/j.eneco.2021.105322 Arvin M. B. Pradhan R. P. Nair M. S. (2021). Are There Links between Institutional Quality, Government Expenditure, Tax Revenue and Economic Growth? Evidence from Low-Income and Lower Middle-Income Countries. Econ. Anal. Pol. 70, 468–489. 10.1016/j.eap.2021.03.011 Awaworyi Churchill S. Inekwe J. Ivanovski K. Smyth R. (2021). Transport Infrastructure and CO2 Emissions in the OECD over the Long Run. Transportation Res. D: Transport Environ. 95, 102857. 10.1016/j.trd.2021.102857 Bergantino A. S. Intini M. Percoco M. (2021). New Car Taxation and its Unintended Environmental Consequences. Transportation Res. A: Pol. Pract. 148, 36–48. 10.1016/j.tra.2021.03.026 Cardenete M. A. López-Cabaco R. (2021). Economic and Environmental Impact of the New Mediterranean Rail Corridor in Andalusia: A Dynamic CGE Approach. Transport Policy 102, 25–34. 10.1016/j.tranpol.2020.12.007 Chen J. Li Z. Wang W. Jiang H. (2018). Evaluating Bicycle-Vehicle Conflicts and Delays on Urban Streets with Bike Lane and On-Street Parking. Transportation Lett. 10 (1), 1–11. 10.1080/19427867.2016.1207365 Chishti M. Z. Ahmad M. Rehman A. Khan M. K. (2021). Mitigations Pathways towards Sustainable Development: Assessing the Influence of Fiscal and Monetary Policies on Carbon Emissions in BRICS Economies. J. Clean. Prod. 292, 126035. 10.1016/j.jclepro.2021.126035 Chudik A. Pesaran M. H. (2015). Common Correlated Effects Estimation of Heterogeneous Dynamic Panel Data Models with Weakly Exogenous Regressors. J. Econom. 188 (2), 393–420. 10.1016/j.jeconom.2015.03.007 Cipek M. Pavković D. Krznar M. Kljaić Z. Mlinarić T. J. (2021). Comparative Analysis of Conventional Diesel-Electric and Hypothetical Battery-Electric Heavy Haul Locomotive Operation in Terms of Fuel Savings and Emissions Reduction Potentials. Energy 232, 121097. 10.1016/j.energy.2021.121097 Conti J. Holtberg P. Diefenderfer J. LaRose A. Turnure J. T. Westfall L. (2016). “International Energy Outlook 2016 with Projections to 2040,”. No. DOE/EIA-0484 (2016)) (Washington, DC: USDOE Energy Information Administration). Dumitrescu E.-I. Hurlin C. (2012). Testing for Granger Non-causality in Heterogeneous Panels. Econ. Model. 29 (4), 1450–1460. 10.1016/j.econmod.2012.02.014 Fan W. Li L. Wang F. Li D. (2020). Driving Factors of CO2 Emission Inequality in China: The Role of Government Expenditure. China Econ. Rev. 64, 101545. 10.1016/j.chieco.2020.101545 Figliozzi M. A. (2020). Carbon Emissions Reductions in Last Mile and Grocery Deliveries Utilizing Air and Ground Autonomous Vehicles. Transportation Res. Part D: Transport Environ. 85, 102443. 10.1016/j.trd.2020.102443 Grote M. Williams I. Preston J. Kemp S. (2018). A Practical Model for Predicting Road Traffic Carbon Dioxide Emissions Using Inductive Loop Detector Data. Transportation Res. Part D: Transport Environ. 63, 809–825. 10.1016/j.trd.2018.06.026 Habib Y. Xia E. Hashmi S. H. Ahmed Z. (2021). The Nexus between Road Transport Intensity and Road-Related CO2 Emissions in G20 Countries: an Advanced Panel Estimation. Environ. Sci. Pollut. Res. 28, 58405–58425. 10.1007/s11356-021-14731-7 Hadavi S. Rai H. B. Verlinde S. Huang H. Macharis C. Guns T. (2020). Analyzing Passenger and Freight Vehicle Movements from Automatic-Number Plate Recognition Camera Data. Eur. Transp. Res. Rev. 12, 1–17. 10.1186/s12544-020-00405-x Hao L.-N. Umar M. Khan Z. Ali W. (2021). Green Growth and Low Carbon Emission in G7 Countries: How Critical the Network of Environmental Taxes, Renewable Energy and Human Capital Is. Sci. Total Environ. 752, 141853. 10.1016/j.scitotenv.2020.141853 Harvey L. D. D. (2020). Rethinking Electric Vehicle Subsidies, Rediscovering Energy Efficiency. Energy policy 146, 111760. 10.1016/j.enpol.2020.111760 He P. Sun Y. Niu H. Long C. Li S. (2021). The Long and Short-Term Effects of Environmental Tax on Energy Efficiency: Perspective of OECD Energy Tax and Vehicle Traffic Tax. Econ. Model. 97, 307–325. 10.1016/j.econmod.2020.04.003 Heinold A. (2020). Comparing Emission Estimation Models for Rail Freight Transportation. Transportation Res. Part D: Transport Environ. 86, 102468. 10.1016/j.trd.2020.102468 Hu X. Chen N. Wu N. Yin B. (2021). The Potential Impacts of Electric Vehicles on Urban Air Quality in Shanghai City. Sustainability 13 (2), 496. 10.3390/su13020496 Hussain Z. Shahenn W. A. Raza S. H. (2020). Trade, Infrastructure and Geography: an Application of Gravity Model on Asian Economies. Int. J. Transport Econ. 47 (2), 145–169. Hussain Z. Yousaf M. Miao C. (2020). Investigating the Simultaneous Impacts on Transport Infrastructure: Evidence from One-belt One-Road Countries. Int. J. Transport Econ. 47 (4), 401–417. Isik C. Ongan S. Ozdemir D. Ahmad M. Irfan M. Alvarado R. (2021). The Increases and Decreases of the Environment Kuznets Curve (EKC) for 8 OECD Countries. Environ. Sci. Pollut. Res. 28 (22), 28535–28543. 10.1007/s11356-021-12637-y ITF (2019). “Efficiency in Railway Operations and Infrastructure Management,”. ITF Roundtable Reports, 177 (Paris: OECD Publishing). Kejun J. Chenmin H. Songli Z. Pianpian X. Sha C. (2021). Transport Scenarios for China and the Role of Electric Vehicles under Global 2 °C/1.5 °C Targets. Energ. Econ. 103, 105172. 10.1016/j.eneco.2021.105172 Khastar M. Aslani A. Nejati M. (2020). How Does Carbon Tax Affect Social Welfare and Emission Reduction in Finland. Energ. Rep. 6, 736–744. 10.1016/j.egyr.2020.03.001 Kong Q. Shen C. Li R. Wong Z. (2021). High-speed Railway Opening and Urban green Productivity in the post-COVID-19: Evidence from green Finance. Glob. Finance J. 49, 100645. 10.1016/j.gfj.2021.100645 Li G. Zhang R. Masui T. (2021). CGE Modeling with Disaggregated Pollution Treatment Sectors for Assessing China's Environmental Tax Policies. Sci. Total Environ. 761, 143264. 10.1016/j.scitotenv.2020.143264 Lin X. (2020). Multiple Pathways of Transportation Investment to Promote Economic Growth in China: a Structural Equation Modeling Perspective. Transportation Lett. 12 (7), 471–482. 10.1080/19427867.2019.1635780 Ma Q. Murshed M. Khan Z. (2021). The Nexuses between Energy Investments, Technological Innovations, Emission Taxes, and Carbon Emissions in China. Energy Policy 155, 112345. 10.1016/j.enpol.2021.112345 Mariano E. B. Gobbo J. A. Camioto F. C. Rebelatto D. A. N. (2016). CO2 Emissions and Logistics Performance: a Composite index Proposal. J. Clean. Prod. 30, 1e13. Mehmood U. (2021). Contribution of Renewable Energy towards Environmental Quality: The Role of Education to Achieve Sustainable Development Goals in G11 Countries. Renew. Energ. 178, 600–607. 10.1016/j.renene.2021.06.118 Mehmood U. (2021). Transport Energy Consumption and Carbon Emissions: The Role of Urbanization towards Environment in SAARC Region. Integr. Environ. Assess. Manag. 17 (6), 1286–1292. 10.1002/ieam.4463 Mirović V. Kalaš B. Milenković N. (2021). Panel Cointegration Analysis of Total Environmental Taxes and Economic Growth in EU Countries. Econ. Anal. 54 (1), 92–103. Moon H. R. Perron B. (2004). Testing for a Unit Root in Panels with Dynamic Factors. J. Econom. 122 (1), 81–126. 10.1016/j.jeconom.2003.10.020 OECD (2021). Environment at a Glance: Climate Change, Environment at a Glance: Indicators. Available at: www.oecd.org/environment/env-at-a-glance . Oecd K. (2018). OECD Science, Technology and Innovation Outlook 2018. Paris: OECD Publishing. Oryani B. Koo Y. Rezania S. Shafiee A. Khan M. K. Mahdavian S. M. (2021). The Role of Electricity Mix and Transportation Sector in Designing a Green-Growth Strategy in Iran. Energy 233, 121178. 10.1016/j.energy.2021.121178 Pani A. Sahu P. K. Holguín-Veras J. (2021). Examining the Determinants of Freight Transport Emissions Using a Fleet Segmentation Approach. Transportation Res. Part D: Transport Environ. 92, 102726. 10.1016/j.trd.2021.102726 Payus C. M. Vasu Thevan A. T. Sentian J. (2019). Impact of School Traffic on Outdoor Carbon Monoxide Levels. City Environ. Interactions 4, 100032. 10.1016/j.cacint.2020.100032 Peng Z. Wu Q. Li M. (2020). Spatial Characteristics and Influencing Factors of Carbon Emissions from Energy Consumption in China's Transport Sector: An Empirical Analysis Based on Provincial Panel Data. Polish J. Environ. Stud. 29 (1), 217. 10.15244/pjoes/102369 Pesaran M. H. (2007). A Simple Panel Unit Root Test in the Presence of Cross-Section Dependence. J. Appl. Econ. 22 (2), 265–312. 10.1002/jae.951 Pesaran M. H. (2006). Estimation and Inference in Large Heterogeneous Panels with a Multifactor Error Structure. Econometrica 74 (4), 967–1012. 10.1111/j.1468-0262.2006.00692.x Pesaran M. H. (2004). “General Diagnostic Tests for Cross-Sectional Dependence in Panel Data,”. Working Paper. Econ. 0435 (Cambridge University). Petrović P. Lobanov M. M. (2020). The Impact of R&D Expenditures on CO2 Emissions: Evidence from Sixteen OECD Countries. J. Clean. Prod. 248, 119187. Polloni-Silva E. Ferraz D. Camioto F. D. C. Rebelatto D. A. D. N. Moralles H. F. (2021). Environmental Kuznets Curve and the Pollution-Halo/haven Hypotheses: an Investigation in Brazilian Municipalities. Sustainability 13 (8), 4114. 10.3390/su13084114 Polloni-Silva E. Silveira N. Ferraz D. de Mello D. S. Moralles H. F. (2021). The Drivers of Energy-Related CO2 Emissions in Brazil: a Regional Application of the STIRPAT Model. Environ. Sci. Pollut. Res. 28, 51745–51762. 10.1007/s11356-021-14097-w Rafique M. Z. Fareed Z. Ferraz D. Ikram M. Huang S. (2022). Exploring the Heterogenous Impacts of Environmental Taxes on Environmental Footprints: An Empirical Assessment from Developed Economies. Energy 238, 121753. 10.1016/j.energy.2021.121753 Rietmann N. Hügler B. Lieven T. (2020). Forecasting the Trajectory of Electric Vehicle Sales and the Consequences for Worldwide CO2 Emissions. J. Clean. Prod. 261, 121038. 10.1016/j.jclepro.2020.121038 Seum S. Ehrenberger S. Pregger T. (2020). Extended Emission Factors for Future Automotive Propulsion in Germany Considering Fleet Composition, New Technologies and Emissions from Energy Supplies. Atmos. Environ. 233, 117568. 10.1016/j.atmosenv.2020.117568 Shahzad U. Ferraz D. Doğan B. Aparecida do Nascimento Rebelatto D. (2020). Export Product Diversification and CO2 Emissions: Contextual Evidences from Developing and Developed Economies. J. Clean. Prod. 276, 124146. 10.1016/j.jclepro.2020.124146 Shahzad U. Radulescu M. Rahim S. Isik C. Yousaf Z. Ionescu S. A. (2021). Do Environment-Related Policy Instruments and Technologies Facilitate Renewable Energy Generation? Exploring the Contextual Evidence from Developed Economies. Energies 14 (3), 690. 10.3390/en14030690 Sharma I. Chandel M. K. (2020). Will Electric Vehicles (EVs) Be Less Polluting Than Conventional Automobiles under Indian City Conditions. Case Stud. Transport Pol. 8 (4), 1489–1503. 10.1016/j.cstp.2020.10.014 Sohail M. T. Ullah S. Majeed M. T. Usman A. (2021). Pakistan Management of green Transportation and Environmental Pollution: a Nonlinear ARDL Analysis. Environ. Sci. Pollut. Res. Int. 28, 29046–29055. 10.1007/s11356-021-12654-x Song M. Xie Q. Shen Z. (2021). Impact of green Credit on High-Efficiency Utilization of Energy in China Considering Environmental Constraints. Energy Policy 153, 112267. 10.1016/j.enpol.2021.112267 Tan X. Tu T. Gu B. Zeng Y. Huang T. Zhang Q. (2021). Assessing CO2 Emissions from Passenger Transport with the Mixed-Use Development Model in Shenzhen International Low-Carbon City. Land 10 (2), 137. 10.3390/land10020137 Tao R. Umar M. Naseer A. Razi U. (2021). The Dynamic Effect of Eco-Innovation and Environmental Taxes on Carbon Neutrality Target in Emerging Seven (E7) Economies. J. Environ. Manage. 299, 113525. 10.1016/j.jenvman.2021.113525 Tao X. Wu Q. (2021). Energy Consumption and CO2 Emissions in Hinterland Container Transport. J. Clean. Prod. 279, 123394. 10.1016/j.jclepro.2020.123394 Tovar Reaños M. A. (2020). Initial Incidence of Carbon Taxes and Environmental Liability. A Vehicle Ownership Approach. Energy Policy 143, 111579. 10.1016/j.enpol.2020.111579 Umar M. Ji X. Kirikkaleli D. Alola A. A. (2021). The Imperativeness of Environmental Quality in the United States Transportation Sector amidst Biomass-Fossil Energy Consumption and Growth. J. Clean. Prod. 285, 124863. 10.1016/j.jclepro.2020.124863 Verlinghieri E. (2020). Learning from the Grassroots: A Resourcefulness-Based Worldview for Transport Planning. Transportation Res. Part A: Pol. Pract. 133, 364–377. 10.1016/j.tra.2019.07.001 Wang K.-H. Su C.-W. Lobonţ O.-R. Umar M. (2021). Whether Crude Oil Dependence and CO2 Emissions Influence Military Expenditure in Net Oil Importing Countries. Energy Policy 153, 112281. 10.1016/j.enpol.2021.112281 Wang Q. Zhang F. (2020). Does Increasing Investment in Research and Development Promote Economic Growth Decoupling from Carbon Emission Growth? an Empirical Analysis of BRICS Countries. J. Clean. Prod. 252, 119853. 10.1016/j.jclepro.2019.119853 Westerlund J. (2007). Testing for Error Correction in Panel Data. Oxford Bull. Econ. Stats 69 (6), 709–748. 10.1111/j.1468-0084.2007.00477.x Xu B. Sharif A. Shahbaz M. Dong K. (2021). Have Electric Vehicles Effectively Addressed CO2 Emissions? Analysis of Eight Leading Countries Using Quantile-On-Quantile Regression Approach. Sustainable Prod. Consumption 27, 1205–1214. 10.1016/j.spc.2021.03.002 Yan S. de Bruin K. Dennehy E. Curtis J. (2021). Climate Policies for Freight Transport: Energy and Emission Projections through 2050. Transport Policy 107, 11–23. 10.1016/j.tranpol.2021.04.005 Yang T. Liao H. Wei Y.-M. (2020). Local Government Competition on Setting Emission Reduction Goals. Sci. Total Environ. 745, 141002. 10.1016/j.scitotenv.2020.141002 Yuelan P. Akbar M. W. Zia Z. Arshad M. I. (2021). Exploring the Nexus between Tax Revenues, Government Expenditures, and Climate Change: Empirical Evidence from Belt and Road Initiative Countries. Econ. Change Restructuring, 1–31. 10.1007/s10644-021-09349-1 Zeqiraj V. Sohag K. Soytas U. (2020). Stock Market Development and Low-Carbon Economy: The Role of Innovation and Renewable Energy. Energ. Econ. 91, 104908. 10.1016/j.eneco.2020.104908 Zhai M. Huang G. Liu L. Guo Z. Su S. (2021). Segmented Carbon Tax May Significantly Affect the Regional and National Economy and Environment-A CGE-Based Analysis for Guangdong Province. Energy 231, 120958. 10.1016/j.energy.2021.120958 Zhang R. Hanaoka T. (2021). Deployment of Electric Vehicles in China to Meet the Carbon Neutral Target by 2060: Provincial Disparities in Energy Systems, CO2 Emissions, and Cost Effectiveness. Resour. Conservation Recycling 170, 105622. 10.1016/j.resconrec.2021.105622