Do renewable energy and total factor productivity eliminate CO₂ emissions in Turkey?

Abstract

This paper analyzes the impact of renewable energy consumption, total factor productivity (TFP), income, imports, and exports on consumption-based CO₂ emissions in Turkey from 1990 to 2019, utilizing the Autoregressive Distributed Lagged (ARDL) method. The findings indicated that the use of renewable energy, total factor productivity and exports negatively influence CO₂ emissions. On the other hand, the positive impact of income and imports on CO₂ emissions was found. The negative impact of renewable energy consumption and total factor productivity highlights the importance of implementing environmentally friendly measures in the energy sector (particularly raising the proportion of renewable energy in overall energy consumption) and technological innovation.

1 Introduction

Energy is very important in the contemporary world for obtaining the required rate of economic growth. This is due to the fact that energy is a crucial component of all economic activities, without which it would be impossible to produce goods and services. Today, when a country’s gross domestic product (GDP) goes up, so do its energy needs (Sadorsky 2009). As intensive utilization of energy and other natural resources degrades the ecosystem, so do emissions from burning fossil fuels, which raise CO₂ levels and destroy the environment and the atmosphere irreparably (Stern 2006; Mukhtarov et al. 2022a, b; Wang et al. 2023). Consequently, greenhouse gases (nitrous oxide, carbon dioxide, and methane) are leftover parts of economic activity that encourage development and wealth but also contribute to environmental damage (Sokolov-Mladenovićet al. 2016; Mukhtarov et al. 2022b). Additionally, since 1900, global carbon emissions from fossil fuels have risen significantly. CO₂ emissions have grown by approximately 90% since 1970. Nearly 78% of the increase in total greenhouse gas emissions from 1970 to 2011 was contributed by CO₂ emissions from the burning of fossil fuels and industrial operations (Boden et al. 2017).

The unrestricted use of fossil fuels is a worldwide concern that might worsen global warming. Global climate change is a top policy priority for all governments since it affects society’s welfare, economic progress, and the environment. According to the 13th Sustainable Development Goal, the global nature of climate change necessitates the broadest international cooperation aimed at accelerating the reduction of global greenhouse gas emissions and addressing adaptation to climate change's negative effects. This was noted in “Transforming our World: The 2030 Agenda for Sustainable Development”, which was published by the United Nations in September 2015 (UN 2015). According to Nordhaus (1975), a 2 °C increase in global temperature could result in irreversible and insurmountable problems. According to the IPCC (2018), if current trends continue, global warming will reach 1.5 degrees Celsius between 2030 and 2052. Similar estimates and awareness attract academics, so studies that try to limit human-led global warming and reduce its consequences are growing in relevance and popularity (IPCC 2018; Mukhtarov et al. 2022a, b).

In 1997, several developing and industrialized nations adopted the Kyoto Agreement to safeguard the environment and eliminate greenhouse gas emissions. The objective was to reduce emissions of greenhouse gases. Initially, industrialized nations were assigned emission reduction objectives. In subsequent years, however, the amount of greenhouse gas emissions from developing economies has quickly surpassed that of industrialized economies, which accounted for around 50% of the world’s total CO₂ emissions in 2003 (Martínez-Zarzoso and Maruotti 2011; Winkler et al. 2002). In this regard, the essence of sustainable development is lowering global emissions while maintaining high levels of economic growth. There are several ways to minimize CO₂ emissions. One of them is to reduce the use of traditional energy sources and increase the use of renewable energy sources. The increased use of renewable energy (the transition to renewable energy) is one of the top priorities on the agendas of nations throughout the globe. The International Renewable Energy Agency (IRENA) defines energy transition as a plan to switch the world’s energy sector from fossil fuels to carbon-free sources by the middle of this century.

Renewable energy, also known as “clean energy” can be generally defined as energy collected from wave, geothermal, solar, tide, wind, wood, waste, and plant materials (biomass). One of the typical features of the renewable energy transition that makes it essential for countries is that it can help in three critical areas: pollution reduction, energy security, and long-term economic growth. Pollution reduction is one of them, and it receives a lot of attention because of the global importance of the issue. Furthermore, technological progress/innovation and efficiency gains are the two main elements of total factor productivity (TFP), and both offer comprehensive avenues for economic growth as well as a decrease in the consumption of energy (including fossil fuels) and, consequently, CO₂ emissions (see, for example, Huang et al. 2020; Miao 2020; Hasanov et al. 2023). Also, major energy organizations such as the International Energy Agency (IEA) and the International Renewable Energy Agency (IRENA) identify technological progress/innovation as one of the key drivers for reducing pollution (Boshell et al. 2018; Hasanov et al. 2021). Contemporary environmental technology has allowed humans to absorb naturally occurring renewable energy sources and transform them into power or usable heat using equipment such as wind turbines, solar panels, and water turbines, representing that technology has a significant positive influence on environmental protection. Therefore, it is common sense that technological progress is critical for pollution reduction (Suurs and Hekkert 2009; Hasanov et al. 2021).

Also, some recent studies such as Khoshnevis Yazdi and Ghorchi Beygi (2017), Zoundi (2017), Danish et al. (2017), Waheed et al. (2018), Nguyen and Kakinaka (2018), Cheng et al. (2019), Dong et al. (2019), Adams and Acheampong (2019), Mahmood et al. (2019), Akram et al. (2020), Piłatowska et al. (2020), Saidi and Omri (2020), Hasanov et al. (2021), Shahnazi and Dehghan Shabani (2021), Makhdum et al. (2022), and inter alia, revealed that higher renewable energy consumption reduced environmental degradation. In addition, some group studies like Yunfeng and Laike (2010), Apergis et al. (2013), Yin et al. (2015), Jin et al. (2017), Alvarez-Herranz et al. (2017), Li et al. (2017), Zhang et al. (2017), Mensah et al. (2018), Wang et al. (2018, 2019), Huang et al. (2020), Hasanov et al. (2021, 2023) concluded that technological progress causes a decline in CO₂ emissions.

In light of the aforementioned context, the goal of this study is to examine the CO₂ impacts of the use of renewable energy consumption and technological advancement, as well as income and international trade, employing the ARDL method, to propose perspectives for environmental policies in Turkey that would help to reduce carbon emissions. Turkey is a developing nation with a population of around 84.6 million. In 2021, it was the world’s 19th largest economy in terms of GDP, the 79th largest in terms of GDP per capita, and ranked 23rd and 29th in total imports and exports, respectively. The Turkish economy rose 11% in 2021, the highest rate among G20 economies. Because of its large population and higher economic growth, Turkey’s energy consumption (particularly from conventional energy sources) is on the rise, which is considered the main contributor to CO₂ emissions. Additionally, Turkey was ranked among the top 25 countries (markets) in the world based on their attractiveness for renewable energy investment and deployment opportunities using five pillars, that is, macro fundamentals, energy imperative, policy, project delivery, and technology (EY 2022). All the above-mentioned facts make Turkey a particular case for this study.

This study contains the following contributions: first, to the best of our knowledge, this is the first time series research to examine the effect of renewable energy consumption, TFP, income, and international trade on CO₂ emissions in Turkey. Second, the international trade is evaluated using imports and exports rather than trade openness, which doesn’t enable the quantification of export and import effects separately.. Third, we use the new model specification that Hasanov et al. (2021) suggested for empirical estimation, which takes consumption-based CO₂ into account. One advantage of utilizing consumer-based CO₂ as a stand-in for carbon emissions is that it takes both ultimate consumption and foreign trade into account. Since it has been modified to take into account global trade, it is now straightforward to identify carbon emissions that are generated in one economy and consumed in another. Finally, the outcomes of this study set the path for other energy-importing economies.

The remaining sections of the paper are organized as follows: Sect. 2 provides an overview of the previous research. Section 3 presents the model specification briefly and data. Sections 4 introduces the econometric approach. The findings of the econometric estimation and discussion are presented in Sect. 5, and Sect. 8 offers conclusion and policy recommendations.

2 Literature review

Many studies have been conducted to investigate the determinants that influence CO₂ emissions. In this regard, the recent studies investigating the effect of TFP (or technological progress-TP) and renewable energy consumption (RE) on CO₂ emissions in the case of different countries were reviewed.

Khoshnevis Yazdi and Shakouri (2017) examined the long-term influence of renewable energy consumption on CO₂ emissions in the case of 13 European Union (EU) countries, utilizing the panel Dynamic Ordinary Least Squares (DOLS) and Fully Modified Ordinary Least Squares (FMOLS) techniques on the data spanning from 1992 to 2014. The estimation findings revealed that RE has a negative and insignificant effect on CO₂ emissions. It means that RE has a negative influence on CO₂ emissions. Moreover, Khoshnevis Yazdi and Ghorchi Beygi (2017) concluded a negative impact of RE on CO₂ emissions for 25 African economies, applying Pooled Mean Group (PMG) approach to the data between 1985 and 2015. Nguyen and Kakinaka (2019) used panel DOLS and FMOLS techniques to analyze the renewable energy consumption-carbon emissions nexus in the case of 107 economies for the data period 1990–2013. They reached the long-term positive effect from RE to CO₂ emissions in the case of low-income economies, whereas a negative influence for high-income economies. Dong et al. (2019) investigated the impact of renewable energy consumption on CO₂ emissions for 120 economies using data period from 1995 to 2015. The estimation results showed that RE has a negative and insignificant impact on CO₂ emissions. In addition, Piłatowska et al. (2020) revealed that there is a negative and insignificant effect of RE on CO₂ emissions in Spain.

Saidi and Omri (2020) studied the relationship between RE and CO₂ emissions in the case of 15 Organization for Economic Cooperation and Development (OECD) economies. They concluded the long-run negative impact of RE on CO₂ emissions by utilizing Vector Error Correction Method (VECM). Also, Leitão and Lorente (2020) found a negative effect of RE to CO2 emissions for EU-28 countries. Ali and Kirikkaleli (2022) examined the RE influence on consumption-based CO₂ emissions, employing the Nonlinear Autoregressive Distributed Lag (NARDL) method for Italy. The estimation findings concluded that RE decreases CO₂ emissions. Besides, Bilgili et al. (2016) in the case of 17 OECD economies, Bhattacharya et al. (2017) in the case of 85 developed and developing countries, Dogan and Seker (2016) in the case of 15 European economies, Zoundi (2017) in the case of 25 African countries, Danish et al. (2017) for Pakistan, Waheed et al. (2018) in the case of Pakistan, Dong et al. (2018) for 128 countries, Kahia et al. (2019) for 12 Middle East and North Africa (MENA countries), Charfeddine et al. (2019) for MENA countries, Adams and Acheampong (2019) in the case of 46 sub-Saharan African countries, Cheng et al. (2019) in the case of Brazil, Russia Federation, India, Indonesia, China, and South Africa (BRIICS countries), Mahmood et al. (2019) in the case of Pakistan, Akram et al. (2020) in the case of 66 developing countries, Okumus et al. (2021) for G7 countries, Shahnazi and Dehghan Shabani (2021) in the case of EU countries, Balsalobre-Lorente et al. (2021) for 5 European Union countries, and Mukhtarov et al. (2023) for Azerbaijan found the negative impact of renewable energy consumption.

Apergis et al. (2010) analyzed the relationship between CO₂ emissions and renewable energy consumption for 19 developing and developed economies using panel data and found a positive effect of renewable energy consumption on CO₂ emissions. Also, they found unidirectional causality running from renewable energy consumption to CO₂ emissions. In addition, Zeb et al. (2014), Jaforullah and King (2015), and Bélaïd and Youssef (2017) concluded similar findings. On the contrary, unidirectional causality running from CO₂ emissions to RE was revealed by Menyah and Wolde-Rufael (2010), Sebri and Ben-Salha (2014), and Lin and Moubarak (2014). In some studies, such as Apergis and Payne (2014) and Jebli et al. (2016) concluded bidirectional causality.

In the case of Turkey, Pata (2018) investigated the effect of RE on CO2 emissions, employing Autoregressive Distributed Lags, Fully Modified Ordinary Least Squares (FMOLS), and Canonical Cointegrating Regressions (CCR) methods. The result of the estimation indicated that RE has a negative and insignificant impact on CO₂ emissions. Also, Shan et al. (2021) studied the impact of RE on CO₂ emission in Turkey, using the bootstrapping ARDL-bound testing method. They achieved that an increase in RE reduces CO₂ emissions. Raihan and Tuspekova found that a 1% rise in RE resulted in reductions in CO₂ emissions of 0.43% in the case of Turkey. Besides, Karaaslan and Çamkaya (2022) evaluated the CO₂ emissions impact of RE, employing the ARDL technique, and revealed that RE decreased the CO₂ emissions for Turkey. Adebayo et al. (2023) concluded that 1% negative and positive shifts in RE reduce CO₂ emissions by 0.00497%, and 0.9104% for Turkey, respectively.

Additionally, several studies have included RE and TFP (or TP) together in their analysis. For example, Hasanov et al. (2021) investigated the effects of TFP and RE on CO₂ emission in the case of Brazil, Russia, India, China, and South Africa (BRICS economies) employing the Cross-Sectional Augmented Autoregressive Distributive Lag (CS-ARDL) model to the data spanning from 1990 to 2017. Hasanov et al. (2023) also found the negative impact of TFP and RE on CO₂ emission in Azerbaijan. The empirical results concluded that TFP and RE have a negative influence on CO₂ emissions. Adebayo et al. (2022) analyzed the impact of TP and RE on CO₂ emissions in Portugal using data from 1980 to 2019. The empirical analysis indicated that the effects of RE and technological innovation on CO₂ emissions are negative and positive, respectively. Ansari et al. (2022) examined the effect of TP and RE together on CO₂ emissions in the case of 10 carbon-emitter countries. They found that the impact of TP and RE are revealed to be negative and positive, respectively.

In addition to the above-mentioned studies, a number of other studies have evaluated the relationship between technological progress and CO₂ emissions. Gu et al. (2019) analyzed the influence of energy technological progress proxied by the number of patent applications on carbon emissions for China employing the Generalized Method of Moments (GMM) method. The findings of the estimation showed that the interval for elasticity is from 1.067 to 0.806. This implies that a 1% rise in energy technological progress will increase CO₂ emissions by 0.806–1.067%. Also, Yu and Du (2019) found a positive effect of technological innovation expressed by research and development (R&D) innovation on CO₂ emissions in China. The outcomes of the estimation showed that a 1% increase in R&D investment raised CO₂ emissions by 0.1049%. On the other hand, Cheng et al. (2019) found a positive and insignificant impact of the development of patents on CO₂ emissions at all quantile levels for OECD countries employing the panel quantile regression (PQR) method. Additionally, the negative influence of energy technology patents on CO₂ emissions was found by Wang et al. (2012).

Huang et al. (2020) examined the total factor productivity impact on CO₂ emissions in China using data between 2000 and 2016. According to the estimation results of both the fixed effect and the Generalized Method of Moments (GMM), the coefficients of TFP are found to be insignificant and negative. In addition, Altinoz et al. (2020) investigated the impact of TFP on CO₂ emission, employing Panel Vector Autoregression (PVAR) and GMM methods to the data period from 1995 to 2014. They revealed that TFP has a statistically significant and negative influence on CO₂ emissions in the case of the Top 10 emerging market economies. Furthermore, Ben Lahouel et al. (2021) revealed that there is a statistically significant and negative impact from TFP to CO₂ emissions in Tunisia. Zhou et al. (2013) evaluated the impact of TFP for China, applying the SYS-GMM method to the panel data between 1995 and 2009. They reached that TFP itself does not decrease CO2 emissions, but it does so through the optimization and upgrading of industrial structures in China. Besides, Wang et al. (2018) considered R&D expenditure as a measure of technological progress. They revealed that there is a negative effect on CO₂ emissions utilizing the Ordinary Least Squares (OLS) technique. Additionally, Yunfeng and Laike (2010), Apergis et al. (2013), Yin et al. (2015), Jin et al. (2017), Alvarez-Herranz et al. (2017), Li et al. (2017), Zhang et al. (2017), Mensah et al. (2018) and Wang et al. (2019) concluded that technological progress decreases CO₂ emissions. In addition to these studies, the relationship between technological innovation and CO₂ emissions in the case of G8 countries was studied by Abid et al. (2022). The findings of FMOLS indicated that there is a statistically significant negative influence of technological innovation on CO₂ emissions in the long term for the data covering 1990–2019.

Two beneficial insights emerge from the reviewed above-mentioned studies for the goal of this investigation. Firstly, several studies have revealed that there is a negative effect of RE and TFP on CO₂ emissions in different countries. Secondly, none of these studies have considered the CO₂ emissions impact of RE and TFP together alongside income and international trade for Turkey in their estimation. To close this gap, the main goal of this study is to evaluate the impact of RE, TFP, income, export, and import on consumption-based CO₂ emissions in the case of Turkey.

3 The functional specification and data

This study’s functional specification was suggested by Hasanov et al. (2021). The detailed information can be found in Hasanov et al. (2021), and we do not extensively discuss it here. Hasanov et al. (2021) used the production function and several assumptions that are common in literature to first develop a connection for energy, where it is a function of the prices of capital and labor, as well as its own price, TFP, and income. Considering the functional specification proposed by Hasanov et al. (2021), the model specification of this study might be stated as follows:

$${\mathrm{CO}}_{2}={a}_{0}+{a}_{1}\mathrm{RE}+{a}_{2}\mathrm{TFP}+{a}_{3}\mathrm{GDP}+{a}_{4}\mathrm{IMP}+{a}_{5}\mathrm{EXP}+e,$$

(1)

where ${\mathrm{CO}}_{2}, \mathrm{RE}, \mathrm{TFP}, \mathrm{GDP}, \mathrm{IMP},\mathrm{ and EXP}$ are the natural logarithmic transformations of consumption-based CO₂ in per capita terms, renewable energy consumption, and total factor productivity, income measured by real GDP per capita, the share of imports in GDP and share of exports in GDP. The $e$ is the error term. Considering the functional specification suggested by Hasanov et al. (2021), we anticipated that real GDP per capita, and share of imports in GDP will have positive effects, while the export share of GDP, renewable energy, and TFP will have negative impact on consumption-based CO₂.

A dependent variable is consumption-based carbon dioxide (CO₂) emissions, in per capita terms. It is measured in million tons of carbon dioxide (MtCO₂). Recent studies suggest that it is better to consider consumption-based CO₂ as opposed to territorial-based CO₂. (Hasanov et al. 2021, 2018; Liddle 2018a, b; Mikayilov et al. 2020). RE stands for the consumption of renewable energy as a share of total final energy usage. It is anticipated that using renewable energy obtained from non-fossil fuel sources would lower consumption-based CO₂. Income (GDP) is expressed by GDP per capita in 2015 US dollars. Total factor productivity (TFP) is indicated by a total number of patents attained by both residents and non-residents. Exports (EXP) are expressed as a proportion of GDP in constant 2015 US dollars. Imports (IMP) are measured as a proportion of GDP in constant 2015 US dollars. Since the model specification proposed by Hasanov et al. (2021), the effect of exports and imports on CO₂ may be assessed separately rather than together. Exports are anticipated to have a negative influence on consumption-based CO₂ because of the quantity of products generated in one country but consumed in another. Due to the quantity of products generated abroad but consumed in the domestic country, imports are anticipated to have a positive influence on consumption-based CO₂. The data of RE, TFP, GDP, IMP, and EXP have been retrieved from the World Bank database (WB 2021). The CO₂ emissions data has been retrieved from Global Carbon Atlas (GCA 2022). All variables are given in the logarithmic form. We used annual data spanning from 1990 to 2019.

4 Econometric methodology

The influence of renewable energy consumption, TFP, income, imports, and exports on consumption-based CO₂ emissions is evaluated by using the ARDL technique in this article. First, the non-stationarity properties of the used variables are checked. For this, the Augmented Dickey–Fuller (Dickey and Fuller 1981, ADF) unit root tests will be utilized.

Second, the Bounds Testing approach to cointegration is applied in order to do the test that determines whether the cointegration link exists. We will finally evaluate the long-run link between the variables using the Bounds Testing Approach to Autoregressive Distributed Lagged (ARDL, Pesaran et al 2001; Pesaran and Shin 1999) model after validating the existence of cointegration between the used variables.

The ARDL is more reliable and has superior performance for usage with small samples, such as the one that was utilized in this research. The ARDL uses dynamic specification, which makes it possible to take into account the influence of lagging values of both the dependent and the independent variables. It can be noted that, this research will focus solely on the long-term relationship between CO₂, RE, GDP, TFP, EXP and IMP. Finally, the Canonical Cointegrating Regression (CCR), the Fully Modified Ordinary Least Squares (FMOLS), and the Vector Error Correction Model (VECM) are utilized for the robustness check.

Due to the fact that similar investigations used the aforementioned methodologies, the detail information about them will not be discussed here. Dickey and Fuller (1981), Phillips and Hansen (1990), Johansen (1990), Park (1992), Johansen (1995), Pesaran and Shin (1999), and Peseran et al. (2001), amongst others, have all contributed to the compilation of this exhaustive information.

5 The results of the empirical analysis and discussion

The ADF unit root test is first used to assess the stationarity features of CO₂, RE, GDP, TFP, EXP, and IMP. Table 1 documents the outcomes of the unit root test. According to the findings of the ADF, all variables become stationary at the first difference but are non-stationary at their levels. Consequently, it is possible to assess the cointegration link between the used variables.

Table 1 The ADF test results

Next, we applied the Bounds Testing Approach to Autoregressive Distributed Lagged to the data in order to conduct a cointegration test and estimate the impact of RE, GDP, TFP, EXP, and IMP on CO₂ emissions. Additionally, the residuals of the model are evaluated for Gauss–Markov conditions, and all the findings are in accordance with the criteria. Additionally, the model was checked for misspecification, and it concluded that there was not a problem with misspecification. The findings of the ARDL method and associated Bounds cointegration test are shown in Table 2.

Table 2 ARDL test results

As can be seen from Table 2, the estimation findings revealed that a 1% increase in RE brings CO₂ emission down by 0.17%, as shown in Table 2. An increase in the use of renewable energy, which is generated from non-fossil fuel sources, would reduce CO₂ emissions. Our results aligned with findings of many studies in the case of developing countries, such as Khoshnevis Yazdi and Ghorchi Beygi (2017), Zoundi (2017), Danish et al. (2017), Waheed et al. (2018), Adams and Acheampong (2019), Cheng et al. (2019), Mahmood et al. (2019), Akram et al. (2020), Hasanov et al. (2021) which concluded that RE has a negative impact on CO₂ emissions. Additionally, the negative and statistically significant effect of RE was revealed by some studies such as, Shan et al. (2021), Raihan and Tuspekova (2022), Karaaslan and Çamkaya (2022), and Adebayo et al. (2023) in the case of Turkey.

Ceteris paribus, CO₂ can be decreased by 0.03% if the TFP is increased by 1% in the long-term. This negative association is aligned with the model specification suggested by Hasanov et al. (2021) that TFP reduces CO₂ emissions. This is a reasonable argument that the two key elements of TFP are technological progress and efficiency gains coming from increased labor skills and capital management, which ought to minimize CO₂ emissions. In addition, our results are in line with several studies like, Yunfeng and Laike (2010), Apergis et al. (2013), Yin et al. (2015), Jin et al. (2017), Alvarez-Herranz et al. (2017), Li et al. (2017), Zhang et al. (2017), Mensah et al. (2018), Wang et al. (2018, 2019), Huang et al. (2020), Hasanov et al. (2021) and Hasanov et al. (2022) concluded that technological progress decreases CO₂ emissions. On the other hand, the positive effect of TFP (or TP) found by some studies like, Gu et al. (2019) for China, Yu and Du (2019) for China, Adebayo et al. (2022) for Portugal, Ansari et al. (2022) for 10 carbon-emitter countries. The difference may be accounted for by country-specific features, the size and frequency of the analyzed periods, the model specification, estimation techniques, and inter alia.

In addition, it is found that income proxied by real GDP per capita has a positive and statistically significant influence on CO₂ emissions. According to this, a 1% rise in real GDP per capita raises CO₂ emissions by 0,19%. Our findings are aligned with economic theory. According to theory, rising output or income is associated with rising consumption of intermediate and final goods and services, which raises CO₂ emissions. Our research findings are aligned with the conclusions of previous studies devoted to developing countries like Mikayilov et al. (2018) and Mukhtarov et al. (2021), Yang et al. (2017), Shuai et al. (2017), Chisti and Sinha (2022), Danish (2019), Hasanov et al. (2019), Hasanov et al. (2021) and Hasanov et al. (2023) which revealed a positive impact of GDP.

Additionally, it is discovered that the influence of import and export are positive and negative, accordingly, which is aligned with predictions and theoretical consequences provided in the section on the theoretical framework. A 1% increase in exports reduces CO₂ emissions by 0.84%, whereas a 1% increase in imports raises CO₂ emissions by 0.43%. One possible reason for these results is that Turkey focuses largely on the import of raw materials/resources (such as crude oil, natural gas, etc.) and export most of production-based products and services (such as, cars, motor vehicles, refined petroleum, etc.) which mostly consumed in abroad. The top three imported goods entering Turkey in 2021 were “mineral fuels, mineral oils, and product of their distillation”. In that year, the import value of these goods was almost 49 billion dollars. As observed in Sect. 3, when a country exports more commodities, it consumes less of them domestically, leading to a decrease in consumption-based CO₂ emissions. Increased imports imply higher local consumption, which boosts consumption-based CO₂ levels. It’s important to note that Hasanov et al. (2021) for the BRICS, Hasanov et al. (2018) for oil-exporting economies, Liddle (2018b) for 20 Asian economies, Liddle (2018a) for 117 countries, Mikayilov et al. (2020) and Hasanov et al. (2023) for Azerbaijan revealed that there is a positive influence from imports to consumption-based CO₂ emissions, while there is a negative influence from exports to consumption-based CO₂ emissions.

Ultimately, the numerical values of the long-term impact of RE, GDP, TFP, EXP, and IMP on CO₂ emissions were evaluated, employing VECM, CCR, and FMOLS as a further robustness check. The estimation results from all three different methods are given in Table 3, in Appendix. As can be seen from Table 3, estimation results of the employed methods supported the findings of ARDL.

6 Conclusion and policy recommendations

The research examines the impact of renewable energy consumption, TFP, income, imports and exports on consumption-based CO₂ emissions. As a first step, the variables were examined to determine whether they have a unit root problem. The results validated the variables’ stationarity in the first differenced form, which enable us to test the variables to see if they have a common long-run trend. Second, utilizing the Bound test was assessed the cointegration relationship between renewable energy consumption, TFP, income, exports, imports, and CO₂ emissions for Turkey. In the end, the existence of the long-term relationship between these variables were assessed by using the ARDL method. The findings of ARDL indicated that, over the long run, the RE, TFP and EXP cause a decline in CO₂ emissions. A 1% increase in RE, TFP, and EXP leading to a decrease in CO₂ emissions by 0.17%, 0.03%, and 0.84%, respectively. On the contrary, empirical research found that a higher import level and a higher GDP per capita both contributed to an increase in CO₂ emissions.

The results of this study may provide important information to decision-makers in the development of technology, energy use, and trade-related policies for CO₂ emissions. This study’s conclusions may help policymakers to understand the role of technological progress and renewable energy use in CO₂ emission reduction. It is general knowledge that technical advancement and increased usage of renewable energy are ecologically harmless. Therefore, authorities in Turkey should keep encouraging the growth of these elements. Particularly, increasing renewable energy investments is a strategy that Turkey should prioritize in tackling its carbon emission problem. Currently, Turkey has some practices with the aim of developing renewable energy projects. In this context, Turkey makes legal arrangements to encourage and facilitate renewable energy investments. Especially thanks to many new regulations enacted in recent years, some facilities are provided for renewable energy projects. On the other hand, Turkey offers various mechanisms and incentives to promote renewable energy projects. In this context, significant premiums are paid to investors within the scope of the Renewable Energy Source Electricity Production Support Mechanism (EPDK 2023). This makes renewable energy projects much more attractive in the eyes of investors. Similarly, Turkey offers some tax advantages for renewable energy investments. In this context, these investors are exempt from certain taxes. This situation offers a significant cost advantage to renewable energy investors.

As stated above, Turkey is currently aiming to increase renewable energy projects with some applications. However, Turkey needs to take some new steps to increase its renewable energy projects in the future. To increase these investments in the short term, government incentives should be increased. The biggest disadvantage of renewable energy projects is the high costs. This situation causes the interest of investors to decrease in these projects. Therefore, government incentives should be sustained to reduce the costs of these projects as a short-term solution. However, this is not a very sustainable practice in the long run. In other words, providing government incentives for the development of renewable energy projects is not a practice that can be sustained for a very long time. In this context, increasing technological investments as a long-term solution plays a very important role. Thanks to the application of new technologies, reducing the costs in renewable energy projects will be possible. This will attract the attention of investors as it will increase the profitability of the projects. In conclusion, using advanced technology and the transition to renewable energy might provide Turkey with an opportunity to obtain sustainable energy and economic developments and lessen its dependency on imported fossil fuel energy sources, such as oil, coal, and natural gas.

Appendix 1

See Table 3.

Table 3 Long-run coefficients from different methods