The interrelationships between carbon dioxide emissions and innovation: international evidence

ABSTRACT

The key concernment for countries in achieving sustainable development goals is reducing emissions. Most studies focused on investigating the role of technological innovation in reducing emissions. The reaction of innovation needs to be considered when increasing emissions. The objective of this research is to examine the interrelationship between technological innovation and carbon emissions for 71 nations in the world from 1996 to 2020, applying simultaneous equation modelling (SEM) with three-stage least squares (3SLS). Outcomes demonstrate a bidirectional relationship between innovation and carbon emissions. Specifically, innovation influences emissions negatively, while carbon emissions influence innovation positively. Renewable energy and forest areas negatively influence emissions, while tourism and agriculture positively influence emissions. Trade and human capital positively impact innovation. Institutional quality enhances environmental quality, while it negatively influences innovation. Several policies promote innovation and reduce emissions towards sustainable development goals.

KEYWORDS:

• Technology innovation
• CO₂ emissions
• SEM
• 3SLS
• institutional quality

Abbreviations

SEM	=	Simultaneous equation model
3SLS	=	Three-stage least squares
REC	=	Renewable energy consumption
CO2E	=	Carbon dioxide emissions
kt	=	Kilotons
TI	=	Technological innovation
HC	=	Human capital
GTI	=	Green technology innovation
EGP	=	Economic growth pressure
AEPI	=	Alternative energy production innovation

1. Introduction

Carbon dioxide emission (CO₂E) levels are the main driver of pollution and global warming.¹ According to statistics from the World Bank, since 1990 global CO₂E has increased by about 60 per cent. The total amount of global CO₂E increased from 21,284042.79 kilotons (kt) in 1990 to 33,566427.59 kilotons in 2020. CO₂E is the main component of greenhouse gases, which have a major impact on global warming. The increasing greenhouse effect leads to a series of social and environmental problems such as melting ice, water shortages, droughts, climate change, and forest fires, which can cause catastrophic impacts on human society. At the 26th Conference on Climate Change (COP₂6) held in Glasgow, countries committed to reducing CO₂ emissions by 45% by 2030 and achieving zero emissions by 2050. Therefore, countries should take effective measures to reduce global CO₂ emissions. According to Weitzman (1997), technological change is of particular importance in describing key environmental problems, mainly long-term and large-scale environmental problems. De Bruyn (1997) believes that technological change plays an important role in reducing environmental pollution, including changes in fuel mix and the use of more energy-efficient production technology. Likewise, Newell and Pizer (2008) show that future technological developments help reduce the intensity of emissions from economic activities and also provide a solution to climate change adaptation. Several papers have explored various options for reducing CO₂ emissions, such as, renewable energy (Raihan 2023), technological innovation (TI) (Khan et al. 2023), forest cover (Raihan and Tuspekova 2022b), and eco-friendly tourism (Raihan and Tuspekova 2022b). TI is one of the effective ways to limit emissions because it promotes the use of clean energy sources to control pollution emissions (Thi, Tran, and Nguyen 2023). TI can help businesses limit their dependence on fossil energy by improving the efficiency of clean energy production (Xin et al. 2021). Moreover, the widespread application of TI in electricity generation may reduce energy consumption and curb CO₂E (Khan et al. 2022).

The world has encountered major challenges in climate change, economic development activities depend on fossil energy sources, leading to increased CO₂E and causing environmental degradation. To solve this problem, technological innovation is necessary to curb CO₂E and reduce environmental degradation. Since 1990, there has been a significantly rising trend in the number of patents related to technological innovation. According to statistics from the World Bank, since 1990, the number of patents related to technological innovation has increased by 30 per cent. The total number of patent applications has risen from 997.500 in 1990 to 3.281.900 in 2020. Many studies show that technological innovation allows us to curb CO₂E (Khan et al. 2023; Thi, Tran, and Nguyen 2023). However, many researchers focus on examining the link between TI and CO₂E from the perspective of examining how the former impacts the latter. The question is whether there exists a reverse effect of CO₂E on the development of TI. That is how the trend in the development of technologies has changed in response to the increase of CO₂E. According to Su and Moaniba (2017), the number of climate-change–associated-innovations is positively responding to rising levels of CO₂E from gas and liquid fuels and CO₂E from solid fuel consumption has a negative impact on innovation. Likely, Wang et al. (2020) examined the impact of CO₂E on eco-innovation. Their outcome indicated that rising CO₂E tends to increase eco-innovation, showing climate change creates opportunities for Chinese innovators. However, there is currently little empirical evidence investigating the impact of CO₂E on innovation, this study will provide further empirical evidence on the impact of CO₂E on innovation by investigating the bidirectional link between technological innovation and CO₂E for 71 countries to assess the role of technological innovation in the fight against climate change. First, we will evaluate the impact of TI on CO₂E. Second, the authors will evaluate the response of innovation to cope with the increase of CO₂E. From the research results, some policy implications will be given for countries aiming for sustainable development.

Besides, the authors also investigate the role of human capital, trade, GDP growth and institutional quality on innovation. Human capital refers to training, education and professional initiatives to enhance the knowledge and skills of employees, which will lead to innovation, promote productivity and sustain competitive advantage (Schultz 1993). Some studies have been carried out in several countries and show human capital as a tool for promoting innovation (Danquah and Amankwah-Amoah 2017; Sarto et al. 2019). According to Grossman and Helpman (1991) and Coe and Helpman (1995), import trade is a means of technology transfer among countries. It is believed that developing nations can achieve technology spill-overs from developed nations by importing intermediate goods, thereby raising the level of TI of developing countries. Also, the role of government is important for green innovation implementation, especially environmental regulations (Borghesi, Cainelli, and Mazzanti 2015; Elgin and Mazhar 2013). A comparative picture of the institutional factors underpinning innovation systems is provided by Polenske (2007). In this study, we will incorporate to evaluate the effect of CO₂E, GDP growth, human capital, trade and institutional quality on innovation.

Some studies have focused on the link among CO₂E and their determinants such as REC, technological innovation (TI), economic expansion, trade openness, tourism, agriculture, forest area, and institutional quality (Raihan 2023; Sadiq et al. 2023; Thi, Tran, and Nguyen 2023; Wei and Lihua 2023; Yuan et al. 2022). According to Thi, Tran, and Nguyen (2023) REC and TI are negatively related to CO₂E, while international tourism and GDP growth show the opposite. Raihan (2023) indicated that REC, agricultural productivity, and forest area have a negative influence on CO₂E. According to Blackman (2010) and Percival et al. (2021) institutional quality, law enforcement effectiveness and environmental standards enforcement are also key factors determining environmental pollution. Yuan et al. (2022) stated that institutional quality has a negative moderating role in the nexus between green innovation and CO₂E. When institutional quality is high, green innovation reduces carbon emissions more strongly. In contrast, Obobisa, Chen, and Mensah (2022) claimed that institutional quality positively impacts CO₂E. In this research, the authors will incorporate to evaluate the role of TI, REC, tourism, agriculture, forest area, GDP growth, and institutional quality in CO₂E.

This paper has several contributions to existing literature in some ways. First, most previous research studies have concentrated on the one-way relation between TI and CO₂E. Many studies have investigated the influence of innovation on CO₂E (Amin et al. 2023; Baba Ali et al. 2023; Su et al. 2023; Zakari, Khan, and Alvarado 2023). Meanwhile, a few researchers have examined the effect of CO₂E on innovation (Wang et al. 2020). Empirical evidence of a bidirectional relationship among them is scanty (Zhao et al. 2023) for China; (Khattak et al. 2020) for BRICS economies. To fill this gap, this paper investigates the bidirectional relation between TI and CO₂E. Our research will put more proof to the literature by investigating this interrelationship in the case of 71 nations worldwide. According to our knowledge, this research is the first effort to examine the bidirectional relation between innovation and CO₂E for world nations. Second, there is a lack of research evaluating the impact of TI, REC, agriculture, tourism, forest area and GDP growth on CO₂E for countries worldwide, our study also intends to fill this gap. Thirdly, very few studies have investigated the effects of CO₂E, GDP growth, trade, and human capital (HC) on technological innovation. Indeed, Shang et al. (2022) only focused on the influence of environmental regulation, and import trade on green TI for 30 provinces in China. Yu et al. (2023) only concentrated the impact of economic growth pressure (EGP) on green technology innovation (GTI) for 285 cities in China. Danquah and Amankwah-Amoah (2017) only focused on the nexus among HC, innovation and technology adoption in sub-Saharan Africa. Wang et al. (2020) only concentrated on the impact of CO₂E on eco-innovation in China. Therefore, our research will put more proof to the literature about the influence of CO₂E, GDP growth, trade and human capital on TI. From the research results, some recommendations will be suggested for policy makers.

This research follows Ren et al. (2021) and Khan et al. (2023) to employ an SEM with the 3SLS method to investigate the bidirectional nexus between innovation and carbon emissions because this technique is more effective than others. The findings indicate a bidirectional relationship between technological innovation and CO₂ emissions in 71 nations. More specifically, TI negatively influences CO₂ emissions, while CO₂ emissions impact innovation positively. This finding supports the view that TI is an important factor in lessening CO₂E. Besides, REC and forest areas have a negative influence on CO₂E, while tourism and agriculture are positively related to CO₂E. Moreover, CO₂E, trade and human capital have a positive impact on TI. Additionally, institutional quality enhances environmental quality by reducing emissions, while institutional quality negative effects on TI. Our results are also confirmed by robustness checks.

The remaining sections of this study are organised as follows: section 2 ‘Literature Review’ highlights significant prior research in this field; the research models and methods will be presented in section 3; section 4 ‘Empirical Results’ section explains the findings; and section 5 ‘Conclusions’ concludes the study with policy implications based on the empirical results.

2. Literature review

Environmental pollution is one of the issues to the world. Achieving the net zero emissions target by 2050 requires countries to come up with many solutions. According to Weitzman (1997), technological change is of particular importance in describing key environmental problems, mainly long-term and large-scale environmental problems. De Bruyn (1997) believes that technological change plays an important role in reducing environmental pollution, including changes in fuel mix and the use of more energy-efficient production technology. Newell and Pizer (2008) show that future technological developments help reduce the intensity of emissions from economic activities and also provide a solution to climate change adaptation.

Many empirical researchers have examined the link between CO₂E and their determinants such as REC, TI, trade openness, economic expansion, agriculture, tourism, and forest area. The research studies of Thi, Tran, and Nguyen (2023) for 53 countries have studied the influence of innovation, tourism, trade openness, and REC on CO₂E for 1990–2019. Their results indicated that REC, TI and trade openness curb CO₂E, while tourism and GDP growth increase CO₂E. Meanwhile, Wei and Lihua (2023) investigated the effect of tourism and innovation on emissions in ASEAN countries for 1995–2018, the results indicated that tourism and eco-innovation have a negative influence on emissions. Likewise, in the case of 35 Belt and Road nations, Khan et al. (2023) studied the relationship between CO₂E, TI, and economic growth for 1985–2019. They show that TI promotes GDP growth and reduces CO₂E, their findings also indicate that school enrolment influencesTI positively, and GDP growth increases CO₂E. Besides, Khurshid et al. (2023) analysed the impact of green innovation on CO₂ emissions in Central-Eastern Europe and found that innovation has a negative influence on emissions. Li et al. (2023) showed that the effect of eco-innovations on carbon emissions is negative by employing the Quantile ARDL approach. Yang et al. (2023) showed alternative energy production innovation (AEPI) is negatively related to CO₂E. Moreover, Hossain et al. (2023) found that renewable and nuclear energy use and innovation lessen carbon emissions by employing the ARDL model, FMOLS and DOLS tests. A negative effect of TI on CO₂E is also found in research studies (Khalid et al. 2023; Shaikh, Taiyyeba, and Khan 2018). Besides, Sadiq et al. (2023) examine the effect of green finance, eco-innovation, REC and carbon taxes on CO₂E for 2001–2020 in BRICS countries using CS ARDL estimation. Outcomes indicated that the influence of these factors on CO₂E is negative. Furthermore, Bhuiyan et al. (2023) studied the influence of green energy development and GDP growth on CO₂E in China for 1980–2020. They show that renewable energy reduces CO₂E, while GDP per capita increases CO₂E. Chandra Voumik et al. (2023) examined the impact of GDP per capita, population, REC, fossil fuels and FDI on CO₂E in Kenya for 1972–2021. Their findings show that the impact of GDP growth on emissions is negative, while population growth is positive with carbon emissions. Besides, renewable energy is beneficial in reducing CO₂E, therefore, to achieve the goals of sustainable development, investment in renewable energy infrastructure is necessary. Wang et al. (2023) studied the influence of REC and non-REC on carbon emissions and GDP growth in 7 Northeast Asian nations for 1970–2020. They indicated that REC increases environmental quality by reducing CO₂E. Ridzuan et al. (2020) investigated the impact of agriculture, REC, and GDP growth on emissions in Malaysia for 1978–2016. Their results indicated that GDP growth and urbanisation raised carbon emissions, while crops, fisheries, and REC significantly decreased emissions. Besides, one of the factors that can reduce environmental degradation is forest area. Waheed et al. (2018) studied the impact of REC, agriculture production and forests on CO₂E in Pakistan for 1990–2014. Their findings show that REC and forests have a negative influence on CO₂E, while the opposite is true for agriculture production. Raihan and Tuspekova (2022a) have studied the influence of GDP growth, energy use, urbanisation, agricultural productivity, and forested area on CO₂E in Kazakhstan during 1996–2020. They stated that GDP growth, energy use, and urbanisation positively affect CO₂E, while agricultural productivity and the forested area have a negative influence on emissions. Raihan (2023) investigated the link between GDP growth, REC, urbanisation, industrialisation, tourism, agricultural productivity, forest area, and CO₂E in the Philippines for 1990–2020. They show that GDP growth, urbanisation, industrialisation, and tourism have a positive impact on emission, while REC, agricultural productivity, and forest area have a negative influence on CO₂E. Raihan et al. (2022) studied the link among CO₂E, GDP growth, REC, urbanisation, industrialisation, TI, and forest area in Bangladesh for 1990–2019. Their results indicate that the positive impact of GDP growth, urbanisation, and industrialisation on CO₂E is found, while REC, TI, and forest area are negatively related to CO₂E. Raihan and Tuspekova (2022b) studied the influence of GDP growth, REC, urbanisation, industrialisation, tourism, agriculture, and forest cover on CO₂E in Turkey for 1990–2020. They stated that growth, urbanisation, industrialisation, and tourism have a positive impact on emissions, while REC, agricultural productivity, and forest area are negatively related to emissions. Raihan and Tuspekova (2022c) analysed the relation among GDP growth, REC, forested area, and CO₂E in Malaysia for 1990–2019. They indicate that economic growth is positively related to emissions, while REC and forested areas have a negative impact on CO₂E, and agricultural production also affects emissions (Raihan and Tuspekova 2022a). Doğan (2019) studied the effect of agriculture on CO₂E in China for 1971–2010 and found that agriculture positively affects CO₂E in China. Besides, Yuan et al. (2022) investigated the effect of green innovation and institutional quality on CO₂E in China and found that green innovation decreased CO₂E. Institutional quality has a negative moderating influence on the nexus between green innovation and CO₂E, when the quality of institutions is high, green innovation strongly curbs CO₂E. Furthermore, Jiang et al. (2022) evaluate the effect of natural resources, energy, quality of institutions, and financial development on CO₂E in the B&R countries from 1995 to 2018. They show that the enhancement of the institutional quality limit CO₂E. The study suggests that countries need to properly address the problems of rapid urbanisation, improve institutions, reduce excessive use of fossil fuels and replace them with renewable energy. Obobisa, Chen, and Mensah (2022) have investigated the effect of green technological innovation and institutional quality on CO₂E in 25 African nations for of 2000–2008. They indicate that green technological innovation and REC significantly reduced CO₂E, while institutional quality, GDP growth, and energy consumption significantly increased CO₂E.

In addition, some studies investigated the response of innovation to the increase of CO₂E. CO₂E from gas and solid fuel use positively effected innovation, while CO₂E from liquid fuel consumption and other greenhouse gas emissions negatively effected innovation (Su and Moaniba 2017). Moreover, Wang et al. (2020) have examined the impact of CO₂E on eco-innovation. Their outcome indicated that rising CO₂E tend to increase eco-innovation, showing climate change creates opportunities for Chinese innovators. Furthermore, Zhao et al. (2023) evaluated the relation between innovation and CO₂E efficiency in China and indicated that there exists a two-way causal relation between innovation and CO₂E efficiency in China. Khattak et al. (2020) have also shown a two-way causality between innovation and CO₂E. Additionally, other studies have examined the link between TI and its determinants such as import trade, human capital, GDP growth and institution quality. For instance, Shang et al. (2022) have studied the influence of environmental regulation, and import trade on GTI for 30 provinces in China from 2008 to 2017. The findings indicate that at first environmental regulation improves GTI and then impedes it, while import trade can stimulate green technology innovation. Furthermore, Dotta and Munyo (2019) have studied the nexus between a country’s innovation and trade openness. They explore that trade openness is positively related to innovation, and research shows that international trade is a natural path to promote innovation. Additionally, Yu et al. (2023) have investigated the impact of EGP on GTI for 285 cities in China for 2006–2018. They found that EGP is negatively related to GTI. Moreover, Shen et al. (2021) have studied the influence of GDP growth target restriction on GTI. They stated that top-down goal setting and amplification of economic growth can hinder green technology innovation. Besides, Danquah and Amankwah-Amoah (2017) have studied the nexus among HC, innovation and technology adoption in sub-Saharan Africa for 1960–2010. They found that human capital positively influences the adoption of technology, while it is insignificant in innovation. Besides, Lee, Wang, and Ho (2020) show that more corruption and the worst governance negatively affect innovations. Likewise, Rodríguez-Pose and Zhang (2020) evaluated the impact of institution quality on a firm`s innovate capacity and found that poor institutional quality is an important barrier for TI. Their outcomes indicate that a weak rule of law, high corruption, and deficient regulatory quality significantly undermine innovation.

A review of existing studies on CO₂E and innovation reveals some research gaps that this study seeks to fill. First, in this study, the authors investigate the two-way relationship between CO₂E and TI using a simultaneous equation model (SEM). This is a gap for this research because through the literature review, the authors noticed that many researchers have investigated either the impact of TI on CO₂E or the impact of CO₂E on TI. According to Zhao et al. (2023), there is a causality relation between innovation and CO₂E. Therefore, the authors will put more proof to the existing literature on the interrelationship between TI and CO₂E in the context of 71 world countries. Secondly, many studies are investigating the influence of REC, TI, agriculture, forest cover and tourism on CO₂E in BRICS countries, in Kenya, 7 Northeast Asian nations, Malaysia, Pakistan, Kazakhstan, the Philippines, Bangladesh, and Turkey. Sadiq et al. (2023) examined the effect of green finance, eco-innovation, REC and carbon taxes on CO₂E in BRICS countries. Raihan (2023) investigated the link between GDP growth, REC, urbanisation, industrialisation, tourism, agricultural productivity, forest area, and CO₂E in the Philippines. However, very few studies evaluate the impact of technological innovation, REC, agriculture, tourism, forest area, GDP growth and institutional quality on CO₂E for 71 countries in the world. Therefore, the authors will put more proof to existing literature on the impact of TI, REC, agriculture, tourism, forest area, GDP growth and institutional quality on CO₂E for such countries. Thirdly, many studies investigate the influence of factors on innovation, such as CO₂E, GDP growth, trade, human capital and institutional quality. Wang et al. (2020) studied whether rising CO₂E accelerates eco-innovation in China. Besides, Zhao et al. (2023) examined the influence of TI on CO₂E efficiency in China. Furthermore, Shang et al. (2022) studied the effect of environmental regulation and import trade on GTI for 30 provinces in China. Yu et al. (2023) investigated the impact of EGP on GTI for 285 cities in China. Shen et al. (2021) studied the impact of GDP growth target restriction on GTI. Danquah and Amankwah-Amoah (2017) studied the nexus among HC, innovation and technology adoption in sub-Saharan Africa. Khan et al. (2023) investigated the link among TI, economic growth and CO₂E in 35 Belt and Road nations. Through a literature review, the authors found that there is currently a lack of research on the impact of CO₂E, economic growth, trade, human capital and institutional quality on technological innovation for countries in the world. Therefore, the authors will put more empirical evidence on the effect of CO₂E, trade, HC, GDP growth and institutional quality on technological innovation for 71 countries in the world.

In general, many researchers have investigated either the impact of TI on CO₂E or the impact of CO₂E on technology innovation. According to Zhao et al. (2023), there is a causality relationship betweem innovation and CO₂ emissions. Therefore, the authors will put more empirical evidence on the interrelationship between TI and CO₂E in the context of 71 countries.

In general, the hypothesis is formulated as follows:

H1. There is a bidirectional relationship between technology innovation and CO₂ emissions.

3. Methodology and data

3.1 Data and variables

The data of all variables were collected from World Development Indicators, World Bank. Our research concentrates on countries worldwide, we collect data for all variables on CO₂ emissions, technological innovation, renewable energy consumption, agriculture, tourism, forest area, GDP growth, trade, HC, and institutional quality in data availability of the World Bank. To evaluate the bidirectional relation between CO₂ emissions and technological innovation, countries had to have available data for at least five consecutive years. After eliminating missing data, the results yield a total of 71 countries and from 1996–2020, where available data related to the variables to be estimated in this study. Detailed data are indicated in Table 1.

Table 1. Variables, measurement, and data sources.

Variables	Abb.	Variables and measurement	Data Source
Carbon dioxide emissions	CO2	Carbon dioxide emissions (kt)	(WDI,2022)
Technological innovation	TI	The total number of patent applications, non-residents and residents	(WDI,2022)
Renewable energy consumption	REC	Renewable energy consumption (% of total final energy consumption)	(WDI,2022)
Agriculture	AGRI	Agriculture, forestry, and fishing, value added (constant 2015 US$)	(WDI,2022)
Forest area	FOREST	Forest area (% of land area)	(WDI,2022)
International tourism	IT	International tourism, number of arrivals	(WDI,2022)
Gross domestic product	GDP	GDP growth (annual %)	(WDI,2022)
Trade	TRADE	Imports of goods and services (BoP, current US$)	(WDI,2022)
Human capital	HC	Labour force participation rate for ages 15–24	(WDI,2022)

Note: WDI = World Development Indicators.⁵

3.2. Theoretical background and methodology

According to De Bruyn (1997), technological change plays a vital role in lessening environmental pollution, including changes in the fuel mix and the use of more energy-efficient production technology. Furthermore, According to Grossman and Krueger (1995), economic growth causes environmental pollution in the early stages, then improves when the country reaches a certain level of income because the application of better technologies can improve environmental degradation. Economic development activities may affect environmental quality such as tourism, agriculture. According to Pigram (1980), tourism can affect environmental quality in three ways: significantly negative, slightly negative and positive. Additionally, tourism is also a major factor causing climate change (Scott, Peeters, and Gössling 2010). Some empirical studies have also shown that tourism has a negative effect on environmental quality (Raihan 2023; Thi, Tran, and Nguyen 2023). Additionally, the agricultural sector also needs to consider its impact on CO₂ emissions more clearly, according to Fao, 21% of total global CO₂E is generated from agricultural production², and is the second sector contributing to environmental degradation. Some empirical studies have indicated that agriculture decreases emissions (Ridzuan et al. 2020), while others show the opposite (Doğan 2019). Besides, energy consumption is one of the main causes of emissions, renewable energy source investment to replace fossil fuels is necessary. Some empirical studies stated that REC allows an increase in the quality of the environment by mitigating CO₂E. To achieve the sustainable development target of zero emissions by 2050, countries must take specific actions to curb CO₂E. Controlling or preventing deforestation is one of the cheapest ways to decrease CO₂E (Stern 2006). With all these concerns, the authors try to evaluate the role of forests in limiting CO₂ emissions in this research.

Besides, Ehrlich and Holdren (1971) proposed the IPAT (Impact by Population, Affluence, and Technology) model to measure the influence of economic activities on the environment. To test the hypothesis more closely, IPAT was reformatted into the STIRPAT (Stochastic Impacts by Regression on Population, Affluence and Technology) model by Dietz and Rosa (1994). The theory to apply for this study is the scalable STIRPAT model. The STIRPAT model is often used to analyse factors affecting environmental degradation (Ren et al. 2021) as follows: (1) ${\mathrm{I}}_{\mathrm{t}}=\mathrm{a}.{\mathrm{P}}_{\mathrm{t}}^{\mathrm{b}}.{\mathrm{A}}_{\mathrm{t}}^{\mathrm{c}}.{\mathrm{T}}_{\mathrm{t}}^{\mathrm{d}}.{ϵ}_t$(1) where I is the level of environmental degradation and a is the intercept term. P is the size of the population, A is the national influence, and T is the progress of technology. b, c, and d indicate elasticities of the environmental influence on P, A, and T, respectively. ϵ_t is a random error term, and subscript t shows the time dimension of the model.

However, the STIRPAT model standard is a non-linear multivariable equation, so it is very hard to figure out the coefficients of a, b, c, d, and ϵ_t. Therefore, all variables in Equation (1) are often converted to logarithmic form for ease of calculation in specific applications. (2) $\mathrm{LnIt}=\mathrm{Lna}+\mathrm{bLnPt}+\mathrm{cLnAt}+\mathrm{dLnTt}+\mathrm{Lnϵt}$(2) where, the definition of variables in Equation 1 is the same as in Equation 2. Previous research has shown that in addition to the initial variables that were set in the model, other variables can affect the environment. Therefore, other factors need to be considered in the model to minimise misjudgement (Gani 2021).

3.2.1. Model extension

To examine the interrelationships between CO₂E and TI for 71 nations, as well as the factors influencing CO₂E and technological advancement. The authors widen the STIRPAT model according to the STIRPAT framework. More specifically, based on the original model, renewable energy, agriculture, tourism, forest area, trade, and human capital are included in the new expansion model.

3.2.2. Carbon emission model

In general, the construction of the CO₂E production function is suitable for the EKC hypothesis (Grossman and Krueger 1995). To a certain extent, the carbon emissions depend on the level of economic activity. The economic expansion causes harmful effects on the environment while technological innovation may reduce environmental degradation (Thi, Tran, and Nguyen 2023). Renewable resources include sunlight, wind, water movement and geothermal heat, which are clean, environment-friendly sources of energy. Therefore, REC can help to reduce harmful impacts on the environment (Wang et al. 2023). In recent times, countries also have been very concerned about emissions from the agricultural sector, and some studies stated agriculture has an impact on CO₂E (Raihan 2023). In addition, the tourism industry is one of the sectors that contributes to economic growth, and it also has an impact on emissions (Raihan and Tuspekova 2022b). Besides, forest area is important in reducing emissions (Raihan and Tuspekova 2022b). Therefore, the authors express CO₂ emissions (I) as a function of technological innovation, economic growth, labour (P), and other factors (X) that impact CO₂E

The environmental pollution equation is presented as follows: (3) ${\mathrm{I}}_{\mathrm{t}}=\mathrm{T}{\mathrm{I}}_{\mathrm{t}}.\mathrm{GD}{\mathrm{P}}_{\mathrm{t}}.{\mathrm{P}}_{\mathrm{t}}.{\mathrm{X}}_{\mathrm{t}}$(3)

3.2.3. The TI model

TI plays a vital role in stimulating growth and minimising environmental degradation (Razzaq, Fatima, and Murshed 2023). Besides, how is the reaction of innovation to the increase in CO₂ emissions is also of interest to some researchers (Su and Moaniba 2017; Wang et al. 2020; Zhao et al. 2023). Furthermore, many empirical studies have investigated the effect of innovation on GDP growth. Whether growth can create motivation for innovation (Shen et al. 2021), as suggested by Shen et al. (2021), GDP growth will also be included in the technological innovation model. Trade also plays a vital role in promoting innovation (Shang et al. 2022). Also, educational attainment plays an important role in stimulating innovation for businesses and countries. According to Danquah and Amankwah-Amoah (2017), human capital has an impact on innovation. The technological innovation equation is presented as follows: (4) $\mathrm{T}{\mathrm{I}}_{\mathrm{t}}=\mathrm{C}{\mathrm{O}}_{2\mathrm{t}}\mathrm{GD}{\mathrm{P}}_{\mathrm{t}}{\mathrm{P}}_{\mathrm{t}}{\mathrm{Y}}_{\mathrm{t}}$(4) Y represents trade and human capital. Our research focuses on the two-way causal relation between TI and CO₂E, as well as the factors that influence CO₂E across countries, can be reported by simultaneous model equations based on the coalescence of equations (3) and (4), as indicated below: (5) $\begin{array}{rl}& \mathrm{C}{\mathrm{O}}_2=\mathrm{GD}{\mathrm{P}}_{\mathrm{t}}{\mathrm{P}}_{\mathrm{t}}\mathrm{T}{\mathrm{I}}_{\mathrm{t}}{\mathrm{X}}_{\mathrm{t}}\\ & \mathrm{T}{\mathrm{I}}_{\mathrm{t}}=\mathrm{C}{\mathrm{O}}_{2\mathrm{t}}\mathrm{GD}{\mathrm{P}}_{\mathrm{t}}{\mathrm{P}}_{\mathrm{t}}{\mathrm{Y}}_{\mathrm{t}}\end{array}$(5)

3.2.4. Model specification

It is important to identify the two-way nexus between TI and CO₂E, the research results will suggest policy-makers promulgate appropriate policies to improve environmental quality and achieve sustainable development goals. According to the above analytical framework, to comprehensively investigate and research the relationship between TI and CO₂E in countries worldwide based on the STIRPAT model, the analysis uses REC, tourism, agriculture, forest area, trade, human capital, and GDP as control variables based on the relevant literature about factors affecting emissions. Energy consumption for economic development has caused a large amount of emissions in the world. Countries need to replace fossil energy with renewable energy sources to achieve sustainable development goals, net zero emissions by 2050. Moreover, tourism is one of the industries that contributes to each country`s economic development, energy consumption for air transport and tourist-related activities also degrades the environment by increasing emissions (Raihan 2023). Furthermore, agriculture plays a vital role in promoting national growth, and agricultural production also affects emissions. Besides, one of the factors that can help reduce environmental degradation is increasing forest area (Raihan 2023). However, currently, there is only a small amount of literature that considers the dependence of REC, agriculture, forest area, and tourism in carbon emission modelling for countries worldwide. Therefore, this study puts REC, tourism, agriculture, forest area, trade, and human capital into the SEM model with the 3SLS estimator. The simultaneous model equations (5) are redefined by employing the terms of logarithmic linearity: (6) $\mathrm{LnCO}2\mathrm{it}=\alpha 0+\alpha 1\mathrm{LnTIit}+\alpha 2\mathrm{RECit}+\alpha 3\mathrm{LnAGRIit}+\alpha 4\mathrm{FORESTit}+\alpha 5\mathrm{LnTOURit}+\alpha 6\mathrm{GDPit}+\mathrm{ϵit}$(6) Technological innovation can be an effective way to limit CO₂ emissions and promote sustainable development, several factors that drive technological innovation, such as trade, human capital, GDP growth, and CO₂ emissions (Danquah and Amankwah-Amoah 2017; Shang et al. 2022). The endogenous growth models of Romer (1990); Grossman and Helpman (1991); Rivera-Batiz and Romer (1991) and Aghion and Howitt (1992) provide association between trade and economic growth. Trade may support technological diffusion and GDP growth by accessing foreign innovations. Bourdieu (2011) describes the link between human capital and innovation as transformational, meaning that different forms of capital can be converted into resources and other forms of economic gain. It is argued that people who are better educated, have more work experience, and spend more time and resources in honing their skills are better able to secure benefits for themselves as well as contribute to the general prosperity of society. Besides, import trade is a means of technology transfer among nations. It is believed that developing nations can achieve technology spill-overs from developed nations by importing intermediate goods, thereby raising the level of TI of developing nations (Coe and Helpman 1995; Grossman and Helpman 1991). Importing nations can achieve spill-over effects related to import trade by learning by doing. This can raise the level of human resources so that innovation activities can be implemented more effectively, ultimately promoting innovation among businesses (Damijan and Kostevc 2015). Based on theories, previous studies and the above discussion, CO₂E, GDP growth, trade and human capital will be included in the analytical framework of the technological innovation model. (7) $\mathrm{LnTIit}=\text{β}0+\text{β}1\mathrm{LnCO}2\mathrm{it}+\text{β}2\mathrm{GDPit}+\text{β}3\mathrm{LnTRADEit}+\text{β}4\mathrm{HCit}+\mathrm{ϵit}$(7) where LnCO₂ is the natural logarithm of carbon dioxide emissions (kt); Technological innovation (LnTI) is the natural logarithm of the total number of patent applications, non-residents and residents; GDP is the growth rate of gross domestic products (%); Trade (Lntrade) is the natural logarithm of imports of goods and services (current US$); Human capital (HC) proxy by labour force participation rate for ages 15–24; Renewable energy consumption (REC) is the share of renewables energy in total final energy consumption (%); Agricultural value (LnAGRI) is the natural logarithm of agriculture, forestry, and fishing, value added (constant 2015 US$). Forest area (FOREST) is the share of forest area in total land area (%). International tourism (LnTOUR) is the natural logarithm of the number of tourists who travel to a country.

$\alpha 0\mathrm{and}\text{ β}0$ are intercept

$\mathrm{ϵit}$ is an error term.

$\alpha 1,\alpha 2,\alpha 3,\alpha 4;\mathrm{and}\text{ β}1,\text{β}2,\text{β}3,\text{β}4$; indicate the coefficients.

i: indicate country (i = 1. ..N)

t: indicate time (t = 1996. . . . 2020)

TI and CO₂E are two endogenous variables in this paper. To solve concurrency between them (the cause of endogeneity), the SEM method is used, Belsley (1988) and Intriligator (1978) stated that 3SLS is more effective than 2SLS because this technique utilises information about the relation between stochastic disturbance terms of structural equations, strengthening the association between the error term. Therefore, the 3SLS estimator is a coalescence of 2SLS and SUR is commonly utilised. Inherited from previous studies like Ren et al. (2021) and Khan et al. (2023), a simultaneous equation model, specifically 3SLS, will be employed in our research.

4. Results and discussions

4.1. The results of baseline models

Table 2 presents the general statistical results of all variables in this study. Table 3 illustrates a positive correlation between CO₂E and TI and no high correlations between independent variables utilised in Equations (1) and (2). However, the concurrent relation between CO₂E and innovation can only be tested by the simultaneous equation model SEM estimator.

Table 2. Description of the variables.

Variable	Obs.	Mean	SD	Min	Max
LnCO2	1,775	11.164	1.778	7.022	16.208
LnTI	1,775	7.490	2.115	1.386	14.248
REC	1,775	22.184	20.173	0	87.23
LnAGRI	1,775	22.727	1.628	18.073	27.722
FOREST	1,775	28.884	19.350	0.045	73.735
LnTOUR	1,775	15.501	1.681	9.472	19.199
GDP	1,775	3.182	3.822	−15.136	19.681
LnTRADE	1,775	6.265	1.461	0	7.451
HC	1,775	45.717	12.858	16.144	83.03

Table 3. Correlations between variables.

Variables	CO2	TI	REC	AGRI	FOREST	TOUR	GDP	TRADE	HC
LnCO2	1
LnTI	0.18	1
REC	−0.43	−0.14	1
LnAGRI	0.82	0.18	−0.09	1
FOREST	0.08	0.05	0.16	0.06	1
LnTOUR	0.56	0.15	−0.35	0.35	0.13	1
GDP	−0.02	−0.04	0.01	−0.01	−0.09	−0.07	1
LnTRADE	−0.02	−0.05	0.02	−0.01	0.07	0.04	−0.03	1
HC	−0.05	−0.02	0.37	0.05	0.13	−0.05	0.02	0.03	1

Before analysing the two – way relation between TI and CO₂E by applying SEM. First, we test cross-sectional dependence within the panel data by applying Pesaran’s CD test. The outcomes of these cross-sectional dependence tests are reported in Table 4, the p-values of all variables are significant at the 1% level, indicating that hypothesis H0 is rejected, and cross-sectional dependence exists in panel data (Table 4). To ensure reliable regression results, it is necessary to test the panel unit root. However, the first-generation unit root tests often ignore cross-sectional dependencies, which can decrease the reliability of outcomes. Therefore, to solve this problem, the second generation root tests will be applied (Pesaran 2007). Table 5 shows the test outcomes of each selected variable.

Table 4. Results of cross-sectional dependence tests by Pesaran CD test.

Variables	Pesaran CD test	Prob-value
LnCO2	26.294***	0.0000
LnTI	4.885***	0.0000
REC	16.369***	0.0000
LnAGRI	119.915***	0.0000
FOREST	25.330***	0.0000
LnTOUR	141.020***	0.0000
GDP	110.191***	0.0000
LnTRADE	5.114***	0.0000
HC	82.927***	0.0000

***p < 0.01.

Table 5. Results of the panel stationarity tests.

Variables	Pesaran CADF test
	Level		1st difference
	Intercept	Intercept & trend	Intercept	Intercept & trend
LnCO2	−2.092***	−2.284	−3.158***	−3.294***
LnTI	−2.313***	−2.688***	−3.683***	−3.789***
REC	−1.092	−2.075	−2.953***	−3.130***
LnAGRI	−2.282***	−2.479*	−3.599***	−3.859***
FOREST	−1.420	−1.858	−2.218***	−2.286
LnTOUR	−2.142***	−2.080	−3.160***	−3.281***
GDP	−2.431***	−2.458*	−3.847***	−3.860***
LnTRADE	−2.189***	−2.392	−3.817***	−3.947***
HC	−1.971**	−2.076	−2.844***	−2.972***

***p < 0.01, **p < 0.05, *p < 0.1.

Table 6 provides evidence of the interrelationship between TI and CO₂E. Specifically, TI negatively affects CO₂E, while CO₂E has a positive impact on TI. Therefore, hypothesis H1 is accepted.

Table 6. The results of models.

Equation (1) of SEM
LnCO2
Constant	11.752*** (0.184)
LnTI	−0.002***(0.000)
REC	−0.038***(0.002)
LnAGRI	0.0008***(0.000)
FOREST	0.001(0.009)
LnTOUR	0.0001***(0.00)
GDP	−0.001 (0.002)
Country fixed effects	Yes
No. Obs	1,775
R^2	0.947
Equation (2) of SEM
LnTI
Constant	−1317.03***(348.912)
LnCO2	136.068***(28.361)
GDP	−0.126 (1.350)
LnTRADE	4.640** (1.983)
HC	6.911***(1.300)
Country fixed effects	Yes
No. Obs	1,775
R^2	0.776

***p < 0.01, **p < 0.05, *p < 0.1.

Equation (1) of Table 6 shows that innovation has a significantly negative influence on carbon emissions, suggesting that innovation lessens emissions. This is supported by previous results (Cai et al. 2023), which explored that innovation curbed CO₂E. Because innovation can realise the efficient use of fossil energy and it can promote the development of renewable energy sources. Furthermore, the coefficient of the REC variable is −0.038, implying that REC influences emissions negatively. This is in line with the findings of Bhuiyan et al. (2023); Chandra Voumik et al. (2023); Wang et al. (2023); Ridzuan et al. (2020), who found that REC is negatively related to CO₂E. This result is relevant because renewable energy is an environment-friendly energy source, combined with technological innovation that allows efficient use of energy resources, which can reduce CO₂E. Conversely, the coefficient of the agriculture variable is 0.0008, indicating that agriculture tends to increase carbon emissions, this finding is in line with the result of Doğan (2019); Balsalobre-Lorente et al. (2019); Yurtkuran (2021); Waheed et al. (2018). This can be explained by the fact that the use of fertilisers, pesticides and the use of fossil energy for agricultural product production can increase CO₂E. According to Fao, 21% of total global CO₂E is generated from agriculture production³, and is the second sector contributing to environmental degradation. Therefore, to achieve sustainable development goals, countries need to have policies to encourage the production and consumption of products with low carbon emissions. Replacing fossil energy using renewable energy in agricultural production needs to be considered. Additionally, the impact of forest area on carbon emissions is positive but it is not statistically significant. Likewise, tourism is positively related to CO₂E, implying that global tourism contributes to environmental degradation. This outcome is similar to the results of Erdoğan et al. (2022); Khan and Ahmad (2021); Razzaq, Fatima, and Murshed (2023). Most of the emissions from tourism activities are generated by the transportation industry. So, it is necessary for innovation related to transportation, renewable energy sources, and building convenient infrastructure to shorten travel distances, save time, save costs, and reduce emissions. Lastly, the coefficient of economic expansion is −0.001 but it is not statistically significant.

Equation (2) of Table 6 shows CO₂E is significantly positive with innovation, implying that increasing CO₂E tends to stimulate innovation. This result is supported by the findings of Su and Moaniba (2017); Wang et al. (2020). The findings indicate that emissions are directly proportional to innovation ability. This result supports the argument that climate change is an opportunity for innovators, climate change is not always a threat (Lee, Min, and Yook 2015; Linnenluecke, Smith, and McKnight 2016). Furthermore, the impact of economic expansion on innovation is negative but it is not statistically significant. While trade is positively related to innovation, showing that trade tends to stimulate innovation, this outcome is similar to the finding of Shang et al. (2022); Dotta and Munyo (2019), who found that import trade enhanced green technological innovation in China during 2008–2017. Because in these countries, technology transfer has been carried out effectively, businesses absorb new technologies through import-export activities and learn through imitation to improve technological innovation capacities. Finally, the coefficient of the human capital (HC) variable is 6.911, indicating that HC tends to promote innovation. This finding is similar to the outcomes of Dakhli and De Clercq (2004); De Winne and Sels (2010); Danquah and Amankwah-Amoah (2017); Munjal and Kundu (2017); Diebolt and Hippe (2019); Sarto et al. (2019), who explored that HC has a positive influence on innovation. Human resources are important assets of nations, and developing human capital to promote innovation is essential for nations. HC is the driving force for innovation, so countries need to invest in the education and training of high-quality human resources.

4.2. Robustness checks

Table 7 shows the outcomes of robustness checks as the authors consider subsamples, the results illustrate that a two-way relation between carbon emissions and innovation still takes for two countries groups, high – upper middle-income countries and low-lower middle-income countries. The impact of TI on CO₂E is negative, while the effect of CO₂E on TI is positive for both two groups. Thess findings are aligned with the outcomes of Cai et al. (2023). Additionally, the impact of the forest area on CO₂E is negative and statistically significant. This result is the same as the research results of Li et al. (2021); Raihan et al. (2022); and Raihan (2023), who explored forest area is negatively related to CO₂E. Because forests play an important role in absorbing CO₂E and preventing erosion, this helps combat climate change. Therefore, countries need to increase forest areas and avoid deforestation and forest burning. Additionally, REC also has a negative impact on CO₂E for both two groups. This finding is aligned with the findings of Thi, Tran, and Nguyen (2023). Likewise, agriculture and tourism have a positive influence on CO₂E for both two groups. These outcomes are supported by the findings of Thi, Tran, and Nguyen (2023) for 53 nations and Doğan (2019). Regarding factors affecting innovation, import trade and human capital affect innovation positively for both two groups. These findings are supported by the outcomes of Shang et al. (2022) and Dakhli and De Clercq (2004). GDP growth impacts innovation negatively but is not statistically significant.

Table 7. Robustness checks with considering subsamples.

Equation (1) of SEM
LnCO2
	High – upper middle-income nations	Low – lower middle-income nations
Constant	12.647***(0.407)	11.177***(0.102)
LnTI	−0.001*** (0.001)	−0.002***(0.000)
REC	−0.024***(0.002)	−0.046***(0.002)
LnAGRI	0.000***(0.000)	0.004***(0.000)
FOREST	−0.023***(0.008)	−0.026*** (0.010)
LnTOUR	0.000***(0.000)	0.000***(0.000)
GDP	−0.000 (0.001)	−0.004 (0.002)
Country fixed effects	Yes	Yes
No. Obs	1,375	400
R^2	0.985	0.979
Equation (2) of SEM
LnTI
	High – upper middle-income nations	Low – lower middle-income nations
Constant	−278.366 (410.405)	−418.148**(203.522)
LnCO2	78.824**(33.420)	39.493***(15.021)
GDP	0.724 (1.154)	−1.002 (0.592)
LnTRADE	5.159*(2.682)	4.942** (2.388)
HC	5.586***(0.992)	2.969**(1.195)
Country fixed effects	Yes	Yes
No. Obs	1,375	400
R^2	0.808	0.641

***p < 0.01, **p < 0.05, *p < 0.1.

According to the suggestion of Polenske (2007) and Blackman (2010), the authors add institutions as a dependent variable to check robustness, institutional quality encompass control of corruption (CORRUPTION), regulatory quality (REGULATORY), rule of law (RULE), voice and accountability (VOICE), government effectiveness (GOVERNMENT), political stability (POLITICAL). These data are gathered from the World Bank database, to avoid multicollinearity problems, the authors include these indicators in a separate model. The obtained outcomes are unchanged, for ease of comprehension, we only interpret the main variables in Table 8.

Table 8. Robustness checks with institutional quality.

Equation (1) of SEM
LnCO2
Constant	11.756*** (0.182)	11.837*** (0.190)	11.777*** (0.179)	11.846*** (0.195)	11.789*** (0.197)	11.808*** (0.190)
LnTI	−0.001*** (0.000)	−0.001*** (0.000)	−0.001*** (0.000)	−0.001*** (0.000)	−0.001*** (0.000)	−0.001*** (0.000)
REC	−0.038*** (0.002)	−0.038*** (0.002)	−0.037*** (0.002)	−0.038*** (0.002)	−0.038*** (0.002)	−0.037*** (0.002)
LnAGRI	0.000*** (0.000)	0.000*** (0.000)	0.000*** (0.000)	0.000*** (0.000)	0.000*** (0.000)	0.000*** (0.000)
FOREST	0.000 (0.009)	0.002 (0.009)	0.001 (0.008)	0.002 (0.009)	0.001 (0.010)	0.001 (0.009)
LnTOUR	0.000*** (0.000)	0.000*** (0.000)	0.000*** (0.000)	0.000*** (0.000)	0.000*** (0.000)	0.000*** (0.000)
GDP	−0.001 (0.002)	−0.001 (0.002)	−0.001 (0.002)	−0.001 (0.002)	−0.001 (0.002)	−0.000 (0.002)
CORRUPTION	−0.000 (0.000)
REGULATORY		−0.000*** (0.000)
RULE			−0.000** (0.000)
VOICE				−0.000** (0.000)
GOVERNMENT					−0.000 (0.000)
POLITICAL						−0.000** (0.000)
Country fixed effects	Yes	Yes	Yes	Yes	Yes	Yes
No. Obs	1,775	1,775	1,775	1,775	1,775	1,775
R^2	0.948	0.946	0.952	0.946	0.946	0.948
Equation (2) of SEM
LnTI
Constant	−1285.175*** (349.540)	−1173.814*** (349.996)	−1252.986*** (360.901)	−1226.422*** (348.639)	−1270.222*** (348.984)	−1244.491*** (359.479)
LnCO2	133.876*** (28.384)	127.385*** (28.355)	131.337*** (29.167)	131.916*** (28.281)	133.846*** (28.334)	130.919*** (29.052)
GDP	−0.187 (1.351)	−0.256 (1.345)	−0.189 (1.351)	−0.370 (1.346)	−0.210 (1.349)	−0.037 (1.352)
LnTRADE	4.658** (1.981)	4.743** (1.993)	4.744** (2.026)	4.619** (1.976)	4.555** (1.966)	4.634** (1.984)
HC	7.076*** (1.312)	6.682*** (1.295)	6.693*** (1.299)	6.713*** (1.296)	6.786*** (1.300)	6.883*** (1.300)
CORRUPTION	−0.022 (0.023)
REGULATORY		−0.077*** (0.023)
RULE			−0.019 (0.024)
VOICE				−0.088*** (0.024)
GOVERNMENT					−0.041* (0.023)
POLITICAL						−0.022 (0.021)
Country fixed effects	Yes	Yes	Yes	Yes	Yes	Yes
No. Obs	1,775	1,775	1,775	1,775	1,775	1,775
R^2	0.776	0.778	0.776	0.778	0.777	0.776

***p < 0.01, **p < 0.05, *p < 0.1.

Equations (1) and (2) of Table 8 show innovation negatively influences carbon emissions, while CO₂ emissions positively affect innovation. Table 8 shows a bidirectional relationship between innovation and carbon emissions in all models. Similar to Haldar and Sethi (2021)`s finding, governance indicators reduce CO₂E, implying that government regulations and laws on renewable energy consumption to protect the environment are effective. Countries need to have policies to invest in clean energy projects, encouraging the use of renewable energy in economic activities and daily activities for sustainable development in the future. Likewise, Yuan et al. (2022) and Jiang et al. (2022) found that institutional quality curbed carbon emissions. Contrary to the research results of Obobisa, Chen, and Mensah (2022) for African countries, they explored that institutional quality positively affected CO₂E. However, institutional quality has a significantly inhibitory influence on TI. This is in line with the findings of Rodríguez-Pose and Zhang (2020); Lee, Wang, and Ho (2020); (My Thi Thi and Tran Phu Do 2024) who explored that institutional quality is a barrier for innovation. It seems that because of pressure on growth targets, some policies and regulations have not created opportunities for innovation.

4.3. Discussions

Our effort is to examine the interrelationship between TI and CO₂E in the world`s 71 countries by applying an SEM estimator combined with 3SLS. Our results illustrate that there exists a two-way causal relation between TI and CO<sub>2</sub>E. Results illustrate a negative influence of TI on CO<sub>2</sub>E. These results are supported by the findings of Khan et al. (2023) for 35 Belt and Road nations as well as those discovered by Thi, Tran, and Nguyen (2023) for 53 countries. The results reaffirm that innovation plays a vital role in lessening environmental degradation and that countries’ policies on economic development accompanied by environmental protection should be considered technological innovation as a beneficial policy tool to limit CO<sub>2</sub>E. Besides, the outcomes indicate that CO<sub>2</sub>E has a positive influence on TI. Su and Moaniba (2017) stated that CO<sub>2</sub>E from gas and solid fuel use positively affected innovation. This suggests that rising emissions present an opportunity for innovation. Therefore, countries need to have policies to encourage individuals and businesses to invent new fuel-saving technologies for the production process. In addition, our outcomes illustrate that REC has a significantly negative impact on CO<sub>2</sub>E, implying that REC is an effective tool to lessen environmental pollution and combat the change of climate. Chandra Voumik et al. (2023)`s outcomes also support our findings, they suggest that national policies should pay attention to investing in renewable energy infrastructure to serve each country's sustainable development goals. Currently, some countries are interested in renewable energy investment: such as wind energy and solar energy … .this seems to be gradually effective in reducing emissions and protecting environmental protection. Countries can issue policies to reduce renewable energy prices and limit fossil energy consumption in industrial parks, businesses and homes. It is necessary to encourage people to use energy-saving appliances and use more renewable energy in the home area. Furthermore, agriculture has a significant positive impact on CO₂E, showing that agricultural production causes emissions and increases environmental pollution. Our outcome is similar to the findings of Doğan (2019), who indicated that agriculture positively affected CO₂E in China. Also, according to Fao, 21% of total global CO₂E is generated from agriculture production⁴, and is the second sector contributing to environmental degradation. However, Raihan and Tuspekova (2022b) were the opposite, they illustrated that agriculture negatively affected CO₂E in Turkey. Therefore, countries need to change from traditional agriculture to smart agriculture production, applying technology in agricultural production. This can help increase productivity and reduce emissions into the environment. Recently, many countries have issued policies to import agricultural products with low carbon emissions. This is a necessary policy to limit carbon emissions in agricultural production. In addition, the government can encourage REC, such as wind energy and solar energy for agricultural production. Additionally, encouraging farmers to apply a circular economy in agricultural production, utilise waste products from livestock, and make organic fertiliser for plants, minimising the amount of chemical fertilisers and pesticides, are also effective ways to reduce emissions. Moreover, global tourism positively affects CO₂E, implying that tourism causes environmental pollution. Our finding is aligned with Raihan (2023), who spotted a positive relationship between tourism and CO₂E in the Philippines. Therefore, to achieve the target of sustainable development, sustainable tourism needs to be established to limit emissions. Tourism is one of the important industries and has a major contribution to the economic growth of each country. Countries can issue policies on green, sustainable tourism, forcing local people as well as tourists to take responsibility for their behaviour related to emissions at tourist destinations. It is necessary to raise people’s awareness about the harmful effects of environmental pollution. Using economic energy and limiting waste are beneficial to the environment. Furthermore, it is necessary to encourage transportation technology, using energy-efficient aircraft, ships and vehicles. Our finding shows a positive relation between trade and TI, implying that trade activities promote innovation. This outcome is aligned with Shang et al. (2022), who stated a positive link between import trade and TI in China. Through import trade, businesses have the opportunity to receive modern technology from multinational companies and developed countries. Businesses may absorb new technologies through import-export activities, and learn through imitation to improve technological innovation capacities. Besides, HC also plays a vital role in promoting innovation. This paper finds that HC has a significantly positive impact on TI, indicating that HC enhances TI. This outcome is the same as the finding of Dakhli and De Clercq (2004), who explored a positive relationshipbetween HC and innovation for 59 countries. They argued that HC is positively correlated with the number of patents. They believe that the higher a country’s educational level, the more it promotes national innovation. Therefore, countries, especially developing countries, need to invest more in education, creating quality human resources, promoting innovation as well as protecting the environment.

5. Conclusion and policy implications

5.1. Conclusion

This study investigates the interrelationships between CO₂E and technology innovation in the world`s 71 nations from 1996 to 2020 by employing a simultaneous equation model (SEM) with 3SLS. The empirical outcomes show a two-way relation between CO₂E and TI, this relationship among them still exists for the sample of high-upper middle-income nations and low –lower middle-income nations. The outcomes show that TI, REC, forest area and institutional quality decrease carbon emission, while agriculture and tourism increase carbon emission. Additionally, findings indicate that CO₂E, import trade and human capital stimulate TI, while institutional quality inhibits innovation.

5.2. Policy implications

From the research outcomes, we suggest a few policies as follows:

Firstly, technological innovation curbs carbon emissions. TI plays a vital role for lessening environmental degradation, and that countries’ policies on economic development accompanied by environmental protection should be considered TI as a beneficial policy tool to limit CO₂E. Therefore, countries need to invest financial resources for technological innovation, producing renewable energy sources, creating energy-saving equipments, encouraging TI in agricultural production, creating low-emission products, environmentally friendly agricultural products. It is also necessary to stimulate innovation in the industry, building ecological industrial parks, replacing fossil energy by renewable energy.

Secondly, an increase in CO₂E stimulates technological innovation, so increasing emissions is also an opportunity for individuals and organisations to improve innovation to cope with CO₂E increase. Therefore, countries should have policies to encourage individuals and businesses to innovate technology in production to improve environmental quality.

Thirdly, renewable energy consumption reduces emissions, so countries need to attract investment for renewable energy projects and issue policies to encourage businesses to invest and use renewable energy sources to replace fossil energy in the future. Countries can issue policies to reduce renewable energy prices as well as limit fossil energy consumption in industrial parks, businesses and homes. It is necessary to encourage people to use energy-saving appliances and use more renewable energy in the home area.

Fourth, forests also play a vital role in mitigating CO₂E, therefore, national governments need to pay attention to forest development, increase investment and have forest conservation policies. The governments can encourage the private sector to participate in forest conservation by developing commercial reforestation areas.

Fifth, agriculture increase environmental degradation by raising CO₂E. Therefore, countries need to change from traditional agriculture to smart agriculture production, applying technology in agricultural production. This can help increase productivity and reduce emissions into the environment. The governments should have issued policies to import agricultural products with low carbon emissions. This is a necessary policy to limit carbon emissions in agricultural production. In addition, governments can motivate REC, such as wind energy and solar energy for agricultural production. Additionally, encouraging farmers to apply a circular economy in agricultural production, utilise waste products from livestock, and make organic fertiliser for plants, minimising the amount of chemical fertilisers and pesticides, are also effective ways to reduce emissions.

Sixth, global tourism stimulates CO₂E. Therefore, to achieve the target of sustainable development, sustainable tourism needs to be established to limit emissions. Countries can issue policies on green, sustainable tourism, forcing local people as well as tourists to take responsibility for their behaviour related to emissions at tourist destinations. It is necessary to raise people’s awareness about the harmful effects of environmental pollution. Using economic energy and limiting waste are beneficial to the environment. Furthermore, it is necessary to encourage transportation technology, using energy-efficient aircraft, ships and vehicles.

Seventh, trade and human capital tend to stimulate innovation, so the government needs to pay more attention to education and improving labour qualifications. Besides, encouraging businesses to technology transfer by learning technology from attracting investment, and learning from import-export businesses.

Last but not least, institutional quality has a negative influence on emissions and innovation, thereby showing that national regulations are effective in reducing emissions, but it has not encouraged innovation. Therefore, countries need to pay more attention and issue policies to encourage innovation in the future.

Although this paper has found empirical evidence about the two – way link between technology innovation and CO₂E in the world`s 71 countries. The impact of TI on CO₂E is negative, while the effect of CO₂E on TI is positive. Furthermore, this research also provides empirical evidence about the influence of REC, TI, agriculture, forest area, GDP growth, and tourism on CO₂E and the effect of CO₂E, GDP growth, import trade and human capital on TI. However, This study still has some limitations. First, due to data limitations, this study only covers 71 countries, therefore, further studies can be carried out for more nations. Secondly, future research can examine the influence of financial risk and human capital on CO₂E. Assessing bidirectional causal relationships among variables should also be considered for future research. Finally, future studies may examine the link between financial development, research and development and technology innovation because financial development and research and development also play a vital role in promoting innovation. Assessing the interrelationship between them should also be considered for further research.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This research is funded by the University of Economics and Law, Vietnam National University Ho Chi Minh City/VNU-HCM. This research is a part of a doctoral thesis conducted by Duyen My Thi Thi.

Notes

1 https://www.nationalgeographic.com/environment/article/greenhouse-gases.

2 https://www.fao.org/3/i6030e/i6030e.pdf.

3 https://www.fao.org/3/i6030e/i6030e.pdf.

4 https://www.fao.org/3/i6030e/i6030e.pdf.

5 http://data.worldbank.org/indicator/NY.GDP.PCAP.CD.