Introduction

Carbon sequestration has become a crucial environmental welfare-enhancing agenda across the world since carbon dioxide (CO2) is regarded as one of the vital greenhouse gases which are emitted into the atmosphere to jeopardize global environmental well-being (Balsalobre-Lorente et al. 2021; Ahmed et al. 2021a; Rehman et al. 2021). Hence, to proactively limit CO2 emissions for tackling the emissions-induced climate change problems, almost all of the global nations have committed under the Paris Accord to curb their respective greenhouse gas emission levels, especially emphasizing on the abatement of CO2 emissions (Murshed et al. 2021a, b; Khan et al. 2021). On the other hand, the world economies are also obligated to comply with their Sustainable Development Goals (SDG) commitments regarding the attainment of environmental sustainability alongside economic and social well-being (Rauf et al. 2018; Murshed 2021a). Therefore, carbon sequestration can be considered both a national and international environmental development agenda which require efficient measures to be undertaken at the earliest. However, despite committing to implement relevant CO2-inhibiting policies, the global CO2 emission figures could not be contained; rather the emissions have amplified with time (World Bank 2021).

It is well-established in the literature that greenhouse gas emissions are mainly driven by energy production and consumption-related activities (Al-Taal et al. 2021; Razzaq et al. 2021). The world economies have traditionally boosted their energy employment levels to trade-off environmental deterioration for higher economic gains (Al Mamun et al. 2014; Salahuddin et al. 2016; Destek and Aslan 2017). Consequently, the combustion of fossil fuels to meet the global energy demand has resulted in the persistent release of CO2 and other greenhouse gases into the atmosphere (Baloch et al. 2021; Murshed 2021b, c). According to the International Energy Association (IEA), the global CO2 emission figures reached an all-time high of more than 33.1 Gt in 2018 (IEA 2019). Such a substantial surge in the emission levels is said to be primarily driven by the extraction and utilization of fossil fuels, especially within the power sector. Hence, it is of utmost importance to mitigate energy production-related emissions worldwide. This is relevant because in 2018 the global annual electricity demand rose by 4% which, although triggered the use of renewables and nuclear energy, led to greater utilization of coal and natural gas for generating electricity (IEA 2019). As a result, the use of these fossil fuels drove up the aggregate volume of energy production-based CO2 emissions stemming from the fossil fuel-fired power plants operating worldwide by almost 2.5% (IEA 2019). Under such circumstances, for curbing these energy use-related CO2 emissions, intensifying the use of renewable sources for generating electricity is recommended within the environmental economics narrative (Murshed 2020a, b; Murshed et al. 2020a; Qin et al. 2021; Zeraibi et al. 2021; Nathaniel et al. 2021a).

Controlling the energy production-related CO2 emissions is relatively more important and difficult for the fossil fuel-dependent emerging market economies such as Argentina. This particular Latin American nation has traditionally traded off poor environmental quality for boosting economic growth rate. Hence, in quest of securing greater economic returns, this nation had profusely burnt fossil fuels to generate electricity; consequently, the associated CO2 emission figures of Argentina have steadily gone up with time (World Bank 2021). The greenhouse gas emissions per capita level of Argentina is greater than the average level of the Group of Twenty (G20) countries. Besides, after the agriculture sector, energy is said to be the second-largest driver of CO2 emissions in Argentina. The energy-related CO2 emission problems of Argentina can be clearly understood from the rise in the nation’s fossil fuel dependency for electricity generation purposes, especially in the contemporary era. Figure 1 illustrates that although the shares of fossil and non-fossil fuels in the total electricity output had somewhat converged during 2002, the differences between these shares have amplified since then. In 2015, the share of fossil fuels in Argentina’s aggregate electricity output was twice that of the non-fossil fuel share. Hence, these trends explicitly depict the rise in the nation’s dependency on fossil fuels for power generation purposes, in particular.

Fig. 1
figure 1

The trends in the shares of fossil and non-fossil fuels in electricity output

Full size image

Moreover, the worsening of the fossil fuel dependency within Argentina’s power sector can further be understood from the trends in the average shares of different primary energy sources in the total electricity output figures of Argentina in each decade between 1971 and 2015. Although moving from the 1970s to the 1980s made way for Argentina’s electricity sector to reduce the reliance on primary fossil fuel supplies, the vast fossil fuel dependency of the nation’s power sector has persistently increased from the 1990s onwards. In the 1980s, the average share of renewables in the total electricity output of Argentina was close to 40%. However, this share has steadily reduced in the subsequent decades, declining by more than 10 percentage points by the end of the mid-2010s. On the other hand, among the shares of the fossil fuel sources, it can be seen that in the 1970s, petroleum/oil accounted for more than half of Argentina’s total electricity output. However, from then onwards, this share declined while the share of natural gas simultaneously went up. This was primarily due to the oil price shocks in the international markets which nudged the government to subsidize the price at which natural gas was supplied to the power sector (Climate Transparency 2018). Besides, the government’s decision to impose a tax on CO2 emissions generated from all types of fossil fuel uses apart from natural gas was equally responsible for the rise in Argentina’s natural gas-fired electricity output shares (Climate Transparency 2018). Similar consequences were faced in other emerging market nations such as Bangladesh (Murshed 2021c).

Furthermore, Fig. 2 also portrays the energy production-related CO2 emission woes of Argentina. It illustrates the trends in the nation’s CO2 emissions generated from electricity and heat production (CO2EHP) figures. The fitted trendline shown in Fig. 2 certifies that the CO2EHP levels of Argentina have on average increased by around 3.5% year-on-year between 1971 and 2018. Hence, these trends clearly indicate the aggravation of environmental quality in Argentina whereby proactive measure to reverse these surging trends need to be undertaken keeping into consideration the country’s national and international environmental sustainability-related commitments. Accordingly, the outcomes from this study can be expected to help Argentina formulate relevant policies in this regard.

Fig. 2
figure 2

Trends in CO2 emissions generated from electricity and heat production

Full size image

Apart from fossil fuel dependency within the power sector, it is acknowledged within the environmental economics narrative that globalization also plays a major role in influencing environmental quality (Kihombo et al. 2021; Ahmed and Le 2021; Ahmed et al. 2021b). For instance, as the world economies economically globalize their respective economy by participating in international trade, the fossil fuel-dependent trading partners tend to specialize in the production of pollution-intensive goods and services while the non-fossil fuel-dependent ones are likely to specialize in the production of cleaner commodities (Ahmed et al. 2019; Banerjee and Murshed 2020; Murshed and Dao 2020). Besides, it has also been recognized in the literature that countries with stringent environmental laws can control international trade-driven CO2 emissions; however, this cannot be ensured within countries in which the environmental laws are relatively more relaxed (Murshed et al. 2021c; Nathaniel et al. 2021b). Similarly, economic globalization in terms of bilateral and multilateral financial flows, such as foreign direct investment (FDI), is also believed to influence the CO2 emission figures. In this regard, inflows of dirty FDI can be expected to boost the CO2 emission levels in the FDI-hosting nations (Solarin et al. 2017; Khan and Ozturk 2020). Conversely, clean FDI inflows, through a technological spillover impact on environmental quality, are likely to curb the emission levels of the host nations (Liu and Xu 2021; Ahmad et al. 2021). In the context of Argentina, the nation’s aggravating CO2 emission trends alongside international trade and foreign financial flows implicate that the economic globalization policies implemented by the government have compromised the nation’s environmental well-being. Besides, since Argentina is an emerging market economy, it is likely to globalize its economy further in the years to come. Therefore, if the globalization policies are not aligned with the environmental protection agenda, it can further boost the nation’s energy-related CO2 emission figures.

Against this milieu, this study primarily examines the environmental effects of renewable electricity generation and economic globalization in the context of Argentina between 1971 and 2016.Footnote 1 As opposed to the conventional approaches, this study measures the environmental effects from the production-side channel by quantifying environmental well-being in terms of the changes in the level of CO2EHP in Argentina. From the point of view of Argentina, this study is important because the current government has recently declared the national objective of becoming carbon–neutral by 2050. Besides, Argentina has recently received Green Climate Fund from the World Bank to implement environmental protection policies, especially to curb the energy use-related greenhouse gas emissions. On the other hand, the government has also enacted the Renewable Energy Act and the Renewable Energy Distributed General Law to amplify the volume of renewable electricity produced in Argentina. These initiatives are envisioned to lessen the nation’s dependency on fossil fuels for electricity generation purposes (Climate Transparency 2018). Moreover, Argentina is also a signatory under the Paris Climate Accord whereby the nation has committed to reduce its energy use-related emissions for tackling climate change adversities. Furthermore, the nation is also focused to achieve environmental sustainability as a part of its commitments under the SDG declarations of the United Nations. Hence, the conclusions from this study can be expected to help the government of Argentina to adopt and implement viable renewable electricity transition and economic globalization policies in order to collectively inhibit energy production-related CO2 emissions.

This study contributes to the existing stock of knowledge in several forms. Firstly, the majority of the existing studies examining the effects of macroeconomic factors on CO2 emissions, both in the context of Argentina and other global economies, have focused on the impacts on the total CO2 emission figures (Al-Mulali et al. 2015; Ozturk 2017). In contrast, this is the only study that emphasizes the determinants of CO2EHP in the context of Argentina. The decision to conduct the analysis at such a disaggregated level is motivated from the point of view that after the agriculture sector, energy is the second-largest contributor to Argentina’s total CO2 emission figures. Hence, it is important to evaluate the factors responsible for the persistent rise in Argentina’s CO2EHP levels. Moreover, since the IEA has reported that the bulk of the global CO2 emission figures are accounted for within the energy sector (IEA 2019), especially for electricity production, it is highly important to identify the determinants of CO2EHP. Accordingly, this study specifically assesses the impacts of renewable electricity transition within the power sector and economic globalization on Argentina’s CO2EHP figures.

Secondly, the preceding studies have conventionally scrutinized the impacts of shocks to renewable energy consumption and output levels on Argentina’s total CO2 emission figures (Yuping et al. 2021). However, measuring renewable energy use in terms of the absolute volume does not reflect the actual renewable electricity transition scenario. This is because simply increasing the level of renewable energy use within the power sector is not adequate to facilitate the electricity transition phenomenon unless the shares of renewables in the total electricity output levels are increased (Farhani and Shahbaz 2014). Besides, a rise in this share is also an indication of a decline in fossil fuel dependency. Furthermore, from the perspective of the power sector, a higher renewable energy use share in the aggregate electricity output is synonymous with a greater capacity to generate renewable electricity. Hence, this study addresses this gap in the literature by examining the effects of changes in the renewable electricity output shares on Argentina’s CO2EHP figures.

Thirdly, emphasizing globalization as a whole may not be appropriate to explore the dynamic relationship between globalization and energy production-related CO2 emissions. Rather, it is more relevant to scrutinize the effects of economic globalization activities on such emissions since economic globalization encompasses international trade and FDI inflows, in particular; both these components of economic globalization-related activities are acknowledged to influence the energy production and consumption levels; thus, economic globalization can also be assumed to influence CO2 emissions generated from energy production. Although the existing studies in the context of Argentina have used the overall globalization index to investigate the globalization-CO2 emissions nexus (Yuping et al. 2021), very little is known regarding the impacts on the emissions associated with economic globalization. Lastly, the analysis conducted in this study takes into account the issues of multiple structural breaks in the data. The previous studies have either ignored the structural break concerns or have only considered a single break within the analysis (Okere et al. 2021). However, in the context of Argentina, accounting for multiple structural breaks is important given the nation has endured multiple episodes of macroeconomic shocks during the period of analysis considered in this study.

The subsequent sections present the following: the relevant literature review is presented in the “ “Review of literature” section while the “Empirical model and estimation strategy” section demonstrates the estimation strategy. The findings are reported and discussed in the “Findings and discussions” section. Lastly, the conclusion and policy suggestions are put forward in the “Conclusion” section.

Review of literature

The effects of macroeconomic variables on CO2EHP are yet to be extensively documented in the existing studies. Hence, in this section, we summarize the empirical studies that have investigated the impacts of economic growth, renewable electricity, globalization, and urbanization on total CO2 emissions.

The nexus between economic growth and CO2 emissions

The effects of economic growth on CO2 emissions have mostly been conducted under the theoretical underpinnings of the Environmental Kuznets Curve (EKC) hypothesis (Shahbaz and Sinha 2019; Shahbaz et al. 2019; Sharif et al. 2020). This hypothesis concludes that the nexus between economic growth and CO2 emissions is inverse U-shaped (Sinha and Bhattacharya 2017; Murshed and Dao 2020; Rahim et al. 2021; Koondhar et al. 2021). However, there are certain limitations of this study. For instance, it is not guaranteed that the CO2 emissions would decline beyond a threshold economic growth level; thus, the EKC hypothesis does not hold for all cases (Apergis and Ozturk 2015; Murshed et al. 2020b). This is likely to happen because it may not be possible for a developing country to reduce its fossil fuel dependency and undergo renewable energy transition due to multifaced constraints (Murshed et al. 2021d, e; Murshed and Alam 2021). As a result, the energy production-based CO2 emissions are not likely to decline. Besides, it is also asserted that the validity of the EKC hypothesis is conditional on the impacts of other key macroeconomic factors since environmental pollution is not solely dependent on economic growth. In a study by Salahuddin et al. (2018), the authors employed the Autoregressive Distributed Lag (ARDL) and Vector Error Correction Model (VECM) techniques to check the relationships between economic growth and CO2 emissions in the context of Kuwait. Based on the findings, the authors asserted that economic growth is a factor responsible for the aggravation of the CO2 emission figures of Kuwait both in the short and long run. Besides, the causality estimates supported the economic growth-led CO2 emissions phenomenon without the feedback causation. Accordingly, the authors stressed that Kuwait should invest in carbon sequestration initiatives to counter the trade-off between economic growth and environmental pollution. Using similar estimation techniques in the context of Nigeria, Rafindadi (2016) economic growth, despite lowering energy demand, contributes to higher CO2 emissions. However, the causality results provided evidence of bidirectional causality between these variables to highlight the interdependency. In the context of Pakistan, Ahmed and Long (2013) used the ARDL model and found statistical evidence of the EKC hypothesis holding only in the long run but not in the short run. The authors claimed that in the short run it is not possible for Pakistan to let go of the trade-off between economic growth and poor environmental quality.

Using annual data from 1970 to 2012, the economic growth-CO2 emission nexus in the cases of Brazil, China, India, and Indonesia was explored by Alam et al. (2016). The authors used the ARDL method for both the linear and non-linear specifications. The results from the linear model revealed that economic growth boosts CO2 emissions for the cases of India and China in the long run and India, Indonesia, and China in the short run; in contrast, economic growth could not explain the variations in Brazil’s CO2 emission figures. On the other hand, the findings from the non-linear model showed that the EKC hypothesis hold for Indonesia, China, and Brazil both in the short and long run but not for India. Ahmad et al. (2017) used quarterly frequency data from 1992Q1 to 2011Q1 and employed the ARDL and VECM methods to scrutinize the effects of economic growth on Croatia’s CO2 emission levels. The results verified the authenticity of the EKC hypothesis for Croatia only in the long run while the causality findings highlighted that economic growth causally influences the CO2 emission figures but not the other way round. In another study on Azerbaijan, Mikayilov et al. (2018) used the ARDL, dynamic ordinary least squares (DOLS), fully modified ordinary least squares (FMOLS), and canonical cointegrating regression techniques and concluded that economic growth monotonically boosts CO2 emissions whereby the EKC hypothesis does not hold for Azerbaijan.

The impacts of economic growth on CO2 emissions were also explored using panel data estimation methods. In the context of European Union (EU) member countries, Kasman and Duman (2015) concluded that economic growth does not causally influence CO2 emission levels in the short run. Besides, the authors also recorded evidence of the EKC hypothesis holding for the EU nations in the long run. Salahuddin et al. (2016) used annual data of Organization for Economic Cooperation and Development (OECD) countries from 1991 to 2012 and used the panel Pooled Mean Group (PMG) technique to evaluate the economic growth-CO2 emissions nexus. The findings revealed that economic growth does not affect the short- and long-run CO2 emission figures. Similarly, in the context of the Emerging Seven (E7) countries, Aydoğan and Vardar (2020) stated that the EKC hypothesis is valid for these emerging nations. Besides, the authors concluded that using clean energy can help to reduce the adverse environmental impacts of economic growth in the long run. In the context of 68 emerging, developed, and the Middle East and North African (MENA) countries, Muhammad (2019) employed the seemingly unrelated regression and system generalized method of moments methods and found higher energy consumption leads to higher economic growth which, in turn, boosts CO2 emissions. In line with these findings, the authors concluded that both economic growth and energy consumption play key roles in stimulating CO2 emissions in the countries of concern. Danish et al. (2019) evaluated the economic growth-CO2 emissions nexus for Brazil, Russia, India, China, and South Africa (BRICS). According to their findings, it was asserted that economic growth along with good governance can eventually counter the trade-off between higher economic growth and environmental degradation to verify the authenticity of the EKC hypothesis. Recently, Sun et al. (2020), using the common correlated effects and augmented mean group methods, also found evidence of the EKC hypothesis holding for a panel of selected OECD and Belt and Road Initiative countries.

In the context of Argentina, the economic growth-CO2 emissions nexus was examined using both time series and panel methods. Yuping et al. (2021) recently explored the effects of economic growth on Argentina’s total CO2 emission figures between 1970 and 2018. The findings, using the ARDL approach, revealed that economic growth initially degrades the environment by boosting CO2 emissions but after a certain point this economic growth-CO2 emissions trade-off is eliminated; thus, the EKC hypothesis was affirmed in that study. Similarly, Adebayo et al. (2021a) also employed the ARDL technique and found economic growth to boost Argentina’s CO2 emission levels both in the short and long run. Besides, the estimates from the DOLS and FMOLS methods also generated homogeneous outcomes. Moreover, the causality results revealed evidence of the economic growth-led CO2 emissions phenomenon to further highlight the detrimental environmental impacts associated with economic growth in Argentina. In another study on the G20 countries including Argentina, Pao and Chen (2019) also found statical evidence of the EKC hypothesis to be valid. Koengkan et al. (2020) evaluated the causal relationships between economic growth and CO2 emissions in the context of five Southern Common Market nations including Argentina, Paraguay, Brazil, Uruguay, and Venezuela. The results provided evidence of bidirectional causal associations between these variables which led to the conclusion that the economic growth policies should be linked with the environmental welfare objectives in order to facilitate environmentally friendly economic growth in these Latin American nations. The adverse environmental effects of economic growth were also highlighted in the study by Hdom (2019) in the context of 10 South American countries including Argentina. The author concluded that higher economic growth boosts CO2 emissions especially due to the fossil fuel dependency of these nations for electricity generation purposes.

The nexus between renewable electricity and CO2 emissions

Although energy consumption, especially fossil fuels, is directly associated with CO2 emissions, not many studies have specifically assessed the impacts of energy use on energy-related CO2 emissions. However, the existing literature is saturated with studies focusing on the effects of energy consumption on overall emissions of CO2 (Ozturk and Acaravci 2016). On the other hand, most of the existing studies have explored the primary energy consumption-CO2 emissions nexus (Ben Jebli et al. 2015; Ali et al. 2019), while a relatively less number of studies have actually emphasized the impacts of electricity consumption/production on CO2 emissions (Bento and Moutinho 2016). Nevertheless, it is important to scrutinize the environmental effects associated with the consumption/production of electricity since electricity is the most important form of energy demanded within any economy. Besides, the fossil fuel dependency of a nation is best understood from the share of its total electricity generated from fossil fuels. Among the preceding studies which have shed light on the relationship between electricity use and CO2 emissions, Rahman (2020) assessed the effects of electricity consumption on CO2 emissions, controlling for economic growth and globalization, in the context of the top 10 electricity consuming nations. The results from the panel FMOLS and DOLS exercises revealed that greater consumption of electricity, generated mostly from fossil fuels, leads to higher CO2 emissions. Besides, electricity consumption and CO2 emissions were evidenced to have bidirectional causal associations. Similarly, Saint Akadiri et al. (2020) and Jiang et al. (2021) concluded that electricity consumption stimulates CO2 emissions in Turkey and China, respectively.

Although in the majority of the cases, aggregate electricity consumption was evidenced to trigger greater CO2 emissions, several studies have highlighted the role of renewable electricity in respect of curbing CO2 emissions (Sinha and Shahbaz 2018). Among these, Bento and Moutinho (2016) used the ARDL model using annual data from 1960 to 2011 and found a rise in the level of per capita renewable electricity production contributes to a reduction in the CO2 emission figures of Italy. Besides, the authors asserted that renewable electricity output is not only a credible means of directly curbing CO2 emissions, but it can also relieve fossil fuel dependency to eliminate the trade-off between economic growth and CO2 emissions. More specifically, the authors concluded that renewable electricity generation plays a key role in validating the EKC hypothesis in the context of Italy. Balsalobre-Lorente et al. (2018), in the context of five EU member nations, also found evidence of renewable electricity consumption helping to mitigate CO2 emissions. Besides, measuring renewable electricity consumption in terms of the level of hydroelectricity consumed in Malaysia, Bello et al. (2018) opined that hydroelectricity consumption helps to mitigate CO2 emissions and other types of environmental adversities while fossil fuel-based electricity consumption boosts CO2 emissions. Moreover, the causality analysis, using the VECM approach, revealed that hydroelectricity consumption influences the CO2 emission figures of Malaysia.

Zhang (2018) used the ARDL model in the context of South Korea and found that non-fossil electricity consumption is detrimental to environmental well-being since a rise in the consumption level of this type of electricity stimulates greater emissions of CO2 into the atmosphere. In contrast, Farhani and Shahbaz (2014) argued that renewable electricity consumption does not guarantee a decline in the level of CO2 emissions. The authors concluded that both non-renewable and renewable electricity consumption boosts the CO2 emission levels of selected MENA countries. On the other hand, from the production side, Jiang et al. (2021) opined that greater production of electricity in China is responsible for the nation’s aggravating CO2 emission trends. Besides, Li et al. (2021a, b) used annual data from 1989 to 2019 in the context of China and found that a rise in the renewable electricity output level is likely to mitigate CO2 emissions and help China achieve its carbon neutrality objective. In the same vein, Usama et al. (2020) also employed the ARDL model in the context of Ethiopia and found evidence of renewable and non-renewable electricity generation being effective in curbing the nation’s CO2 emission levels. Among the very few existing studies on the impacts of renewable electricity on energy-related CO2 emissions, Zambrano-Monserrate et al. (2017) examined the effects of renewable electricity, natural gas, and petroleum consumption on energy consumption-based CO2 emissions in Peru. The results from the ARDL analysis revealed that renewable electricity consumption reduces energy-related CO2 emissions only in the long run. However, natural gas and petroleum consumption were found to reduce both the short- and long-run energy-related CO2 emission levels of Peru.

The nexus between globalization and CO2 emissions

The globalization-CO2 emissions nexus was initially evaluated using the overall globalization index. Liu et al. (2020), in the context of the Group of Seven (G7) countries, found evidence of globalization initially degrading the environment by boosting CO2 emissions but later on improving it by curbing CO2 emissions. In the context of Brazil, Russia, India, and China (BRIC), Pata (2021) used the Fourier ARDL method and concluded that globalization increases the CO2 emissions in Brazil and China only but not in India and Russia. In another study on South Africa, Salahuddin et al. (2019) utilized the ARDL technique and showed that globalization does not trigger CO2 emissions in the short run; conversely, globalization was evidenced to stimulated greater emissions of CO2 in the long run. Alola and Joshua (2020) opined that globalization exerts environmental adversities in high, upper-middle, and low-income countries but improves environmental quality in lower-middle-income countries. Besides, the income group-specific estimates showed that globalization triggers relatively higher CO2 emissions in the cases of high-income countries.

Besides, several preceding studies have also assessed the effects of different types of globalization on CO2 emissions. Among these, Shahbaz et al. (2017) employed the ARDL and VECM methods and found that social, economic, and political globalization help to curb CO2 emissions in China. However, the causality estimates revealed that CO2 emissions causally influence the different forms of globalization without the feedback effects. In the context of selected OECD countries, Yang et al. (2021) recently stated that greater degrees of economic globalization are necessary for mitigating CO2 emissions in the countries of concern. Besides, the authors also highlighted the evidence of unidirectional causal relationships stemming from economic globalization to CO2 emissions. Similarly, Majeed et al. (2021) explored the effects of economic globalization on CO2 emissions in the context of the Gulf Cooperation Council (GCC) countries. The authors employed the cross-sectional augmented ARDL method and found evidence of economic globalization can effectively mitigate CO2 emissions in these countries in the long run. In another study concerning the emerging economies, Ulucak et al. (2020) found evidence of financial globalization reducing CO2 emissions in the long run. Besides, using disaggregated indicators of financial and economic globalization, several existing studies have scrutinized the impacts of FDI inflows and international trade on CO2 emissions. Among these, Essandoh et al. (2020) used data from 52 developed and developing nations and found evidence of FDI inflows boosting CO2 emissions in low-income countries while international trade reduces the CO2 emission figures of high-income countries. Similarly, in the context of developed and developing countries from Europe and the Asia Pacific, respectively, Khan and Ahmad (2021) opined that FDI inflows amplify CO2 emissions while international trade abates CO2 emissions.

In the context of Argentina, recently Yuping et al. (2021) employed the ARDL technique and asserted that globalization, as a whole, helps to reduce CO2 emissions. Besides, the authors also found evidence of globalization exerting contrasting environmental outcomes when interacted with the consumption of alternative energy resources. Similarly, Adebayo et al. (2021a) also used the ARDL technique and concluded that globalization, overall, mitigates CO2 emissions in Argentina. It is to be noted that in both the aforementioned studies on Argentina, the authors do not control for structural break issues in the data and overlook the effects of economic and other types of globalization on Argentina’s CO2 emission levels. Among the relevant panel data studies, Nathaniel et al. (2021b) explored the effects of total globalization on CO2 emissions in 18 Latin American and Caribbean nations including Argentina. The panel findings showed that globalization imposes adverse environmental consequences by boosting CO2 emissions in the long run. Besides, bidirectional causal relationships between globalization and CO2 emissions were also put forward. In another study on Argentina and 17 other Latin American and Caribbean countries, Koengkan et al. (2020) concluded that total, social, economic, and political globalization exert adverse environmental impacts by boosting CO2 emissions. In line with these findings, the authors suggested that the globalization policies of these nations should be made environmentally friendly to phases out the associated environmental adversities.

The nexuses between urbanization and CO2 emissions

The findings from the existing studies regarding the effects of urbanization on CO2 are mixed. Pata (2018), in the context of Turkey between 1974 and 2014, employed the ARDL, FMOLS, and canonical regression methods and found that urbanization degrades the environmental quality by boosting CO2 emissions in the long run. Besides, the authors claimed that the adverse environmental impacts of urbanization could be because Turkey is yet to be developed enough to focus on the implementation of CO2-abating policies. Ali et al. (2019) explored the urbanization-CO2 emissions nexus in Pakistan. The results from the ARDL analysis showed that urbanization degrades environmental quality by amplifying CO2 emissions in the long run. Similarly, Mahmood et al. (2020) employed the non-linear ARDL method in the context of Saudi Arabia and found that urbanization boosts CO2 emissions in the long run. In another study on 55 middle-income countries, Lv and Xu (2019) concluded that urbanization increases CO2 emissions in both the short and long run. Using CO2 emissions per capita level as a measure of environmental quality in 48 African nations, Mignamissi and Djeufack (2021) asserted that urbanization boosts CO2 emissions; however, the impacts are heterogeneous across different quantiles of CO2 emissions. Besides, urbanization was evidenced to be relatively more detrimental to environmental quality in the resource-rich African nations than in the resource-poor African nations. Accordingly, these studies have recommended promoting planned urbanization activities to reduce the level of urbanization-induced CO2 emissions. Although the majority of the existing studies have highlighted the adverse environmental effects of urbanization, Muhammad et al. (2020) asserted that urbanization initially triggers CO2 emissions but later on inhibits CO2 emissions in Belt and Road Initiative (BRI) countries.

As far as the impacts of urbanization on CO2 emissions in the context of Argentina are concerned, Adebayo et al. (2021a) concluded that urbanization does not explain the variations in Argentina’s short- and long-run CO2 emission figures. In another relevant study on 18 Latina American and Caribbean nations including Argentina, Adebayo et al. (2021b) found that urbanization contributes to higher CO2 emissions in the long run. Similarly, Nathaniel et al. (2021b) recently documented evidence of urbanization boosting CO2 emissions in the context of a panel of Argentina and 17 other Latin American and Caribbean countries. In the context of selected developed nations including Argentina, Usman et al. (2021) found that urbanization stimulates CO2 emissions only in the long run.

Empirical model and estimation strategy

Empirical model

Conventionally, the environmental impacts were assessed using the IPAT model (Balsalobre-Lorente et al. 2021). According to this modeling approach, environmental impact (I) is said to be conditional on changes in population (P), level of affluence (A), and stock of technology (T) within an economy. However, later on, addressing the limitations of the IPAT model, York et al. (2003) introduced a stochastic variant of the IPAT model which was termed as the STIRPAT model. In this study, we consider an augmented version of the STIRPAT model by including renewable electricity output shares and economic globalization within the model. Besides, in line with the theoretical underpinnings of the EKC hypothesis, we also convert the linear STIRPAT model into a non-linear format. The double-log baseline model employed can be specified as:

$$\mathbf{Model}\boldsymbol\;\mathbf1:\;\ln{\mathrm{CO}2\mathrm{EHP}}_{\mathrm t}=\varphi_0+{\varphi_1{\mathrm{RELECT}}_{\mathrm t}+\varphi}_2{\mathrm{EGI}}_{\mathrm t}+\varphi_3{\mathrm{lnRGDP}}_{\mathrm t}+\varphi_4{\left(\mathrm{lnRGDP}^2\right)}_{\mathrm t}+\varphi_5{\mathrm{URB}}_{\mathrm t}+\varepsilon_{\mathrm t}$$
(1)

where the subscript t refers to the time period of analysis (1971–2016), \({\varphi }_{0}\) is the intercept parameter, \({\varphi }_{k}(k=1, 2,\dots , 5)\) are the elasticity parameters, and \({\varepsilon }_{t}\) is the error term. The dependent variable lnCO2EHP is the natural logarithm of per capita CO2 emissions stemming from the production of electricity and heat. The CO2EHP figures are estimated by summing up three categories of energy production-related CO2 emissions specified by the IEA.Footnote 2 Among the explanatory variables, RELECT refers to the share of total electricity output generated using renewable sources. This variable is included in the model since it gives indications of the renewable electricity generation capacity of Argentina and also portrays the renewable energy transition phenomenon. Furthermore, a higher value also demonstrates a decline in Argentina’s fossil fuel dependency for electricity generation purposes.

The variable EGI denotes the economic globalization index which is estimated considering the involvement of Argentina in international trade and its FDI inflows, in particular. Higher values of this index are synonymous with a rise in the degree of Argentina’s economic globalization with the other world economies. This is because if this elasticity parameter is positive \(({\varphi }_{2}>0)\) then it can be interpreted as a rise in the energy-related CO2 emissions figures of Argentina following a marginal increase in the nation’s economic globalization index (which would take into account a rise in the volume of FDI inflows in Argentina). Under such circumstances, economic globalization, especially in the form of financial globalization in respect of FDI inflows, can be asserted as a means of environmental adversity. The variables lnRGDP and (lnRGDP)2 refer to the real GDP per capita and its squared term, respectively. This variable is used as a proxy for affluence or economic growth in Argentina. As per the theories concerning the EKC hypothesis, the EKC hypothesis is validated if the predicted signs of the associated elasticity parameters are positive and negative, respectively \(({\varphi }_{3}>0 {and }{{\varphi }}_{4}<0)\). Lastly, the variable URB refers to the urbanization rate which is measured as a percentage of the urban residents in the total population of Argentina.

To further assess the joint impacts of renewable electricity output and economic globalization on Argentina’s CO2EHP figures, we augment the interaction term between these variables in Model (1). The augmented model can be specified as:

$${\mathbf{Model}\boldsymbol\;\mathbf2}:\;\ln{\mathrm{CO}2\mathrm{EHP}}_{\mathrm t}=\varphi_0+{\varphi_1{\mathrm{RELECT}}_{\mathrm t}+\varphi}_2{\mathrm{EGI}}_{\mathrm t}+\beta{(\mathrm{RELECT}\ast\mathrm{EGI})}_{\mathrm t}+\varphi_3{\mathrm{lnRGDP}}_{\mathrm t}+\varphi_4{\left(\mathrm{lnRGDP}^2\right)}_{\mathrm t}+\varphi_5{\mathrm{URB}}_{\mathrm t}+\varepsilon_{\mathrm t}$$
(2)

where the variable (RELECT*EGI) is the interaction term between renewable electricity output share and the economic globalization index. The predicted sign of the elasticity parameter β would indicate the joint impact of these variables on CO2EHP in Argentina. Table 1 provides the definitions, units of measurement, and corresponding data sources of the above-mentioned variables.

Table 1 The definitions, units of measurement, and data sources of the variables
Full size table

Estimation strategy

The econometric analysis is structured into four stages. In the first stage, the unit root properties of the variables are assessed (“The unit root analysis” section) while the second stage involves the prediction of cointegration among the variables (“The cointegration analysis” section). In the third stage, the regression analysis is conducted to estimate the elasticity parameters for both the short and long run (“Regression analysis” section). Lastly, in the fourth stage, the causal relationships between the variables are evaluated (“The causality analysis” section).

The unit root analysis

The unit root analysis helps us to understand whether or not a series converges to its mean value; if it converges, then there are no unit root problems but if they diverge then unit root issues can be expected to exist. The analysis of unit root is pertinent because the presence of unit root in the data, synonymous with the problem of non-stationarity, results in spurious regression outcomes (Murshed 2021a). Besides, the order of stationarity among the variables in the respective model is important to determine the optimal regression method that is to be tapped to predict the elasticity parameters. The traditional first-generation techniques assume that there are no structural breaks in the data. However, this assumption is inappropriate since it is highly likely that an economy endures multiple macroeconomic shocks that can affect the impacts of the explanatory variables on the outcome variables. To account for the limitations of the traditional methods, we consider the second-generation unit root estimation technique of Narayan and Popp (2013). As opposed to the conventional approaches, the Narayan and Popp (2013) estimator predicts the unit root properties of the variables in the presence of a structural break in the data. An advantage of this method over the other second-generation methods, such as the Zivot and Andrews (1992), is that it can account for multiple structural breaks in the data. Controlling for more than one structural break is important since Argentina has endured several episodes of macroeconomic shocks, both at national and international levels. According to Narayan and Popp (2013), the unit root outcomes generated by utilizing this technique have more stability and are of the correct size. This method offers two different estimation models: (a) Model A in which the structural breaks are assumed to be located in the intercept; (b) Model B in which the locations of the structural breaks are assumed to be in the intercept and slope of the trend function (Murshed 2021a). Both these models predict test statistics under the null hypothesis of non-stationarity of the respective variable (series). Once the unit root properties are ascertained, the cointegration analysis is conducted.

The cointegration analysis

The existence of cointegrating relationships among the variables included within a model denotes that in the long run, these variables move together. It is important to evaluate the cointegrating relationships because it is meaningless to predict the long-run elasticities in the absence of cointegrating relationships amid the variables of concern. Hence, the cointegration analysis is a necessary condition for performing the long-run regression analysis. In this study, two different cointegration methods are applied. At first, the ARDL Bounds test of Pesaran and Shin (1998) and Pesaran et al. (2001) is applied. This method predicts a joint F-statistic from the respective model under the null hypothesis on non-cointegration among the variables of concern. The unrestricted ARDL model for conducting the Bounds test can be expressed as:

$${\Delta Y}_{{t}}={\alpha }_{{o}}+\sum \nolimits_{{i}=1}^{{n}}{{\alpha }}_{1{i}}{\Delta Y}_{{t}-{i}}+\sum \nolimits _{{i}=1}^{{n}}{\beta }_{1{i},{j}}{\Delta X}_{{j},{t}-{i}}+{\uptheta }_{1}{Y}_{{t}-1}+{\theta }_{2{j}}{X}_{{j},{ t}-1}+{\varepsilon }_{{t}}$$
(3)

where Y is the dependent variable, Xj is a vector of independent variables, and ∆ is the first difference operator. From Eq. (5), the joint significant F-test statistic is predicted under the null and alternative hypotheses shown below:

$$\mathrm{Null\;Hypothesis}: {\theta }_{1}={\theta }_{2}=0\;\;\; (\mathrm{non\;cointegration})$$
$$\mathrm{Alternative\;Hypothesis}: {\theta }_{1}={\theta }_{2}\ne 0\;\;\; (\mathrm{cointegration})$$

However, the ARDL Bounds test does not account for the multiple structural breaks issues in the data. To resolve this limitation, we also employ the Maki (2012) cointegration method which predicts the cointegrating relationships among the variables of concern and identifies up to a maximum of five structural breaks in the data. Under this method, there are four model specifications according to the assumption regarding the location of the structural breaks and trend specifications. These models include (a) Model A which assumes the structural breaks to be in the intercept without a trend assumption; (b) Model B which assumes the structural breaks to be in the intercept with a trend assumption; (c) Model C which assumes the structural breaks to be in the intercept and coefficient without a trend; (d) Model D which assumes the structural breaks to be in the intercept and coefficient with a trend assumption. In the context of this study, the test statistics under the Maki (2012) approach are predicted in respect of Model D. Besides, since the length of the period of the analysis is short, we limit the maximum number of structural breaks to two. Following the cointegration analysis, we conduct the regression analysis to predict the short- and long-run elasticities.

Regression analysis

In this study, we employ the ARDL regression method of Pesaran et al. (2001) to estimate our models. Compared to the other conventionally used time-series regression methods, the ARDL technique has a couple of advantages. Firstly, this method can accommodate variables that are integrated at up to the first difference. Besides, although most of the alternative methods require a common order of integration, the ARDL method considers a mixed order of integration. Secondly, the ARDL approach, as confirmed by the Monte Carlo simulation, is superior and robust to predicting unbiased outcomes for panel data sets of small cross sections and short time dimensions (Ahmad et al. 2017). Thirdly, this technique neutralizes the issues of endogenous covariates and is efficient in handling models with endogeneity concerns. Moreover, even if all explanatory variables are endogenous, the ARDL technique is still able to predict unbiased outcomes (Ahmad et al. 2017). Fourthly, the ARDL method has a provision for assigning different lag lengths for the respective variables included in the model. Lastly, it can predict outcomes for both the short and long run and also provides a speed of adjustment of short-run disequilibrium to long-run equilibrium.

The short- and long-run elasticities are predicted using an unrestricted error-correction model which, in the context of Model 1 specified in Eq. (1), can be expressed as:

$${\Delta \mathrm{lnCO}2\mathrm{EHP}}_{\mathrm{t}}= {\delta }_{0}+ \sum_{\mathrm{f}=1}^{\mathrm{Y}}{\alpha }_{1\mathrm{f}}{\Delta \mathrm{lnCO}2\mathrm{EHP}}_{\mathrm{t}-\mathrm{i}}+ \sum_{\mathrm{f}=1}^{\mathrm{Y}}{\alpha }_{2\mathrm{f}}{\Delta \mathrm{RELECT}}_{\mathrm{t}-\mathrm{i}}+ \sum_{\mathrm{f}=1}^{\mathrm{Y}}{\alpha }_{3\mathrm{f}}{\Delta \mathrm{EGI}}_{\mathrm{t}-\mathrm{i}}+\sum_{\mathrm{f}=1}^{\mathrm{Y}}{\alpha }_{4\mathrm{f}}{\Delta \mathrm{lnRGDP}}_{\mathrm{t}-\mathrm{i}}+ \sum_{\mathrm{f}=1}^{\mathrm{Y}}{\alpha }_{5\mathrm{f}}{\Delta ({\mathrm{lnRGDP}}^{2})}_{\mathrm{t}-\mathrm{i}}+ \sum_{\mathrm{f}=1}^{\mathrm{Y}}{\alpha }_{6\mathrm{f}}{\Delta \mathrm{URB}}_{\mathrm{t}-\mathrm{i}}+ {\beta }_{1}{\mathrm{lnCO}2\mathrm{EHP}}_{\mathrm{t}-1}+{\beta }_{2}{\mathrm{RELECT}}_{\mathrm{t}-1}+{\beta }_{3}{\mathrm{EGI}}_{\mathrm{t}-1}+{\beta }_{4}{\mathrm{lnRGDP}}_{\mathrm{t}-1}+{\beta }_{5}{\left({\mathrm{lnRGDP}}^{2}\right)}_{\mathrm{t}-1}+{\beta }_{6}{\mathrm{URB}}_{\mathrm{t}-1}+{\varepsilon }_{t}$$
(4)

where Y denotes the lag orders which vary across the respective variables and ∆ symbolizes the difference operator. From Eq. (4), the short-run error-correction model can be extracted as follows:

$${\Delta {lnCO}2{HEP}}_{{t}}= {{\varnothing }}_{0}+ \sum_{{f}=1}^{{Y}}{\alpha }_{1{f}}{\Delta {lnCO}2{EHP}}_{{t}-{i}}+ \sum_{{f}=1}^{{Y}}{\alpha }_{2{f}}{\Delta {RELECT}}_{{t}-{i}}+ \sum_{{f}=1}^{{Y}}{\alpha }_{3{f}}{\Delta {EGI}}_{{t}-{i}}+\sum_{{f}=1}^{{Y}}{\alpha }_{4{f}}{\Delta {lnRGDP}}_{{t}-{i}}+ \sum_{{f}=1}^{{Y}}{\alpha }_{5{f}}{\Delta ({{lnRGDP}}^{2})}_{{t}-{i}}+ \sum_{{f}=1}^{{Y}}{\alpha }_{6{f}}{\Delta {URB}}_{{t}-{i}}+{\varnothing {ECT}}_{t-1}+{\varepsilon }_{t}$$
(5)

where ECTt-1 is the one-period lagged error-correction term. The error-correction term shows the speed of adjustment from short-run disequilibrium to long-run equilibrium. The long-run model can be extracted from Eq. (4) can be shown as:

$${\Delta {lnCO}2{EHP}}_{{t}}= {{\varnothing }}_{0}+ \sum_{{f}=1}^{{Y}}{\alpha }_{1{f}}{{lnCO}2{EHP}}_{{t}-{i}}+ \sum_{{f}=1}^{{Y}}{\alpha }_{2{f}}{{RELECT}}_{{t}-{i}}+ \sum_{{f}=1}^{{Y}}{\alpha }_{3{f}}{{EGI}}_{{t}-{i}}+\sum_{{f}=1}^{{Y}}{\alpha }_{4{f}}{{lnRGDP}}_{{t}-{i}}+ \sum_{{f}=1}^{{Y}}{\alpha }_{5{f}}{({{lnRGDP}}^{2})}_{{t}-{i}}+ \sum_{{f}=1}^{{Y}}{\alpha }_{6{f}}{{URB}}_{{t}-{i}}+{\varepsilon }_{t}$$
(6)

The regression analysis is followed by the causality analysis.

The causality analysis

Causality analysis is important to understand the directions of causal associations between a pair of variables. In contrast, the regression analysis simply indicates a marginal change in the independent variable causes a change in the dependent variable but does not provide information if a change in the dependent variable would execute a change in the independent variable. The conventional causality estimators such as Granger (1969) and Toda and Yamamoto (1995a, b) are efficient in predicting the causal properties in the context of large samples. However, these methods do not perform well in case of small samples (Koçak and Şarkgüneşi 2018). To resolve the limitations of these methods, Hacker and Hatemi-J (2006) developed a modified Wald test for estimating the causal relationships. The Monte Carlo simulations showed that the modified Wald test of Hacker and Hatemi-J (2006) uses a leveraged bootstrapped approach to reduce the size distortions which take place in the context of the traditional methods. Hence, the Hacker and Hatemi-J (2006) method are more efficient in handling small samples. Later on, Hacker and Hatemi-J (2012) modified the Hacker and Hatemi-J (2006) method by endogenizing the optimal lag length based on the Schwarz Information Criterion (SIC) and Akaike Information Criterion (AIC). Under this method, the modified Wald statistic is predicted under the null hypothesis of the independent variable not Granger causing the dependent variable. If the value to the test statistic is larger than the bootstrapped critical values generated using the Monte Carlo simulation, then the null hypothesis is rejected to affirm causality.

Findings and discussions

The empirical analysis begins with the evaluation of the unit root properties of the variables using the Narayan and Popp (2013) method. The unit root findings are reported in Table 2. It can be witnessed that all the variables apart from lnRGDP/(lnRGDP)2 and EGI are stationary at the first difference. The statistical significance of the predicted test statistics certifies this claim. Based on these findings, it can be inferred that the variables in both models 1 and 2 are of mixed order of integration, up to a maximum of the first difference, I(1). Besides, the respective locations of the structural breaks are also identified and reported in Table 2. The unit root analysis is followed by the cointegration analysis.

Table 2 Narayan and Popp (2013) unit root analysis
Full size table

The cointegrating relationships among the variables, for the respective model, are firstly assessed using the ARDL Bounds test approach. The corresponding results, as reported in Table 3, show that there are existences of at least one cointegrating equation in both Models 1 and 2. Since the magnitudes of the predicted F-statistics, for the respective model, are above the corresponding 1% and 5% upper bound critical values, the null hypothesis of non-cointegration can be rejected. Hence, it can be asserted that CO2EHP have long-run associations with renewable electricity generation, economic globalization, economic growth, financial development, and urbanization in the case of Argentina. However, since this method does not identify the locations of structural breaks for the respective model, the Maki cointegration (2012) analysis is also performed. The corresponding results, as presented in Table 4, further affirms the presence of cointegrating relationships among the variables in the respective model. Besides, the locations of the two structural breaks for each model are also identified which are included as structural break dummy variables within the models. The cointegration analysis is followed by the ARDL short- and long-run elasticity estimations.

Table 3 The ARDL Bounds test results
Full size table
Table 4 Maki cointegration test
Full size table

The predicted short- and long-run elasticity parameters for Models 1 and 2 are presented in Table 5. It is evidenced that enhancing renewable electricity output shares by 1% leads to a decline in the CO2EHP figures of Argentina by 0.66–0.64% and 0.87–0.99% in the short and long run, respectively. These imply that undergoing renewable electricity transition, by reducing the employment of fossil fuels in Argentina’s power sector, can be effective in mitigating the energy production-based CO2 emission figures. Besides, the relatively larger long-run elasticity parameters imply that continuously increasing the renewable electricity output shares enhances the favorable environmental outcomes by persistently decreasing the associated emissions. These findings are consistent with the results documented in the studies by Bello et al. (2018) and Balsalobre-Lorente et al. (2018) for Malaysia and EU member nations, respectively. Conversely, the elasticity findings reveal that economic globalization is detrimental to environmental well-being in the long run. A 1% rise in the economic globalization index is predicted to boost the CO2EHP figures in the long run by 0.16–0.18%; however, no short-run marginal effect of economic globalization on CO2EHP is witnessed. Hence, in line with these findings, it can be asserted that Argentina’s globalization policies, especially those related to international trade and financial globalization, are not aligned with the nation’s goal of transforming its economy into a low-carbon one. Similar findings were reported for the case of China by Shahbaz et al. (2017).

Table 5 The ARDL short- and long-run results
Full size table

Although the above-mentioned findings portray the positive and negative environmental effects associated with renewable electricity generation and economic globalization, respectively, the predicted elasticity parameter attached to the interaction term highlights the joint favorable impact of these variables on the environment. The negative signs of the corresponding elasticity estimates imply that renewable electricity generation helps to reduce the adverse environmental impacts of economic globalization by jointly reducing Argentina’s CO2EHP figures both in the short and long run. This key finding portrays that it is of utmost importance for Argentina to undergo renewable electricity transition since it not only directly curbs the CO2EHP but also indirectly reduces these emissions by greening the nation’s economic globalization activities. For instance, if the local electricity demand associated with international trade and FDI inflows is met by the renewable electricity output, then the energy production-based CO2 emissions can largely be mitigated in the long run.

Besides, the elasticity estimates also verify the validity of the EKC hypothesis, both for the short and long run, in the context of Argentina. The positive and negative signs of the predicted elasticity parameter attached to lnRGDP and its squared term, respectively, suggest that the economic growth-CO2EHP nexus is inverse U-shaped. Therefore, it can be inferred that the persistent growth of the economy of Argentina can eventually help the nation to control the aggravation of its environmental pollution issues by limiting energy production-related CO2 emissions. These findings are parallel to the findings documented by Alam et al. (2016) for Indonesia, China, and Brazil. ON the other hand, the elasticity estimates also indicated that urbanization is dampening the quality of the environment in Argentina. A 1% rise in the rate of urbanization in Argentina is found to boost CO2EHP by 0.69–0.90% in the short run and by 0.80–1.07% in the long run. These findings imply that the urbanization strategies executed in Argentina are not environmentally sustainable. This is because the urban electricity demand in Argentina is predominantly met by the nation’s fossil fuel-based electricity supplies; the nation generates a large portion of its total electricity output by combusting natural gas. The findings corroborate the results found by Ali et al. (2019) for Pakistan and Mahmood et al. (2020) for Saudi Arabia.

The values of the adjusted R-squared figures for Models 1 and 2 are estimated at 0.84 and 0.76, respectively. Hence, it can be said that around 76–84% of the total variations in Argentina’s CO2EHP figures can be explained by changes in the nation’s renewable electricity output share, economic globalization index, per capita economic growth level, and urbanization rate. The estimated lagged ECT for Models 1 and 2 are evidenced to be negative and statistically significant as well. Accordingly, it can be asserted that any short-run disequilibrium is restored to the long-run equilibrium level at a speed of 84% for the case of Model 1 and by 75% for the case of Model 2. Besides, the diagnostic test findings portray that both the models considered in this study are free from serial correlation, model misspecification, and heteroscedasticity concerns while the variables are also found to be normally distributed. Furthermore, the CUSUM and CUSUMSQ plots, shown in Fig. 3, confirm the stability of the parameters for both models.

Fig. 3
figure 3

The CUSUM and CUSUMSQ plots for Models 1 and 2

Full size image

For a robustness check of the long-run elasticity estimates, the DOLS estimator is employed. Table 6 presents the findings from the DOLS analysis. Overall, the predicted signs of the elasticity parameters predicted using the DOLS method are similar to those predicted using the ARDL method. However, the DOLS elasticity parameters are relatively smaller in magnitude which could be due to the incapability of the DOLS method to accommodate variables with mixed order of integration. Under such circumstances, the ARDL elasticity parameters can be asserted to be relatively more accurate which counters for the under-estimate bias of the DOLS elasticity parameters by accounting for the mixed order of integration among the variables considered in this study. Lastly, the causality analysis is conducted to ascertain the causal associations between the variables.

Table 6 The robustness analysis using the DOLS estimator
Full size table

The causal associations between the variables of concern are scrutinized using the Hacker and Hatemi-J (2012) method. The causality results, as reported in Table 7, provide evidence of unidirectional causalities running from renewable electricity generation, economic globalization, economic growth, and urbanization to the CO2EHP figures of Argentina. Thus, the causality findings provide support to the elasticity estimates reported in Table 5.

Table 7 The Hacker and Hatemi-J (2012) bootstrapped causality analysis
Full size table

Conclusion

The energy sector of Argentina is predominantly reliant on the employment of fossil fuels for producing electricity, in particular. Among the different primary fossil fuels utilized within the power sector, a lion’s share of the total electricity output of Argentina is generated from combusting the indigenous natural gas supplies. As a result, the energy production-related CO2 emission figures of the nation have steadily swollen over the years which, in turn, deteriorated the quality of the environment in Argentina. The rising trends in Argentina’s energy production-related CO2 emissions cast a shadow of doubt regarding the nation’s prospects of complying with the Paris Climate Change Agreement and SDG commitments. Against this backdrop, this study scrutinized the effects of renewable electricity generation, economic globalization, economic growth, and urbanization on the CO2EHP figures of Argentina between 1971 and 2016. The econometric analysis conducted in this study involved the application of the latest methods that control for multiple structural breaks in the data. Overall, the results portrayed that there are long-run associations between CO2EHP and the explanatory variables of concern in the context of Argentina. Besides, the regression analysis showed that enhancing the share of renewable electricity in the total electricity output of Argentina is capable of curbing the energy production-related CO2 emissions both in the short and long run. In contrast, economic globalization and urbanization were evidenced to boost the energy production-based CO2 emission figures. However, the results also confirmed that renewable electricity generation helps to reduce the adverse environmental effects associated with economic globalization. Moreover, the results also verified the authenticity of the EKC hypothesis in the context of Argentina. Furthermore, the causality analysis supported the regression findings by portraying statistical evidence of unidirectional causalities stemming from renewable electricity generation, economic globalization, economic growth, and urbanization to CO2EHP.

In line with these major findings, several policies can be recommended to help Argentina tackle its persistently increasing CO2EHP figures. Firstly, it is of utmost importance for Argentina to transform its power sector by significantly enhancing the renewable electricity output shares. Greater employment of renewable sources, while reducing the use of primary fossil fuels, for electricity generation purposes can be expected to reduce the overall energy production-based CO2 emissions in Argentina. In this regard, Argentina should think of refraining from subsidizing the natural gas supplies to the domestic power plants. Such initiatives can help to control the level of natural gas combusted within these power plants; consequently, the energy production-based CO2 emissions can be minimized. Additionally, investment in the development of renewable energy technologies can also help to facilitate renewable electricity transition within Argentina’s power sector which, in turn, can further reduce the nation’s energy-related emission figures. Secondly, Argentina should also sustainably revisit its economic globalization strategies so that globalization does not boost the energy production-related CO2 emission figures of the nation. Accordingly, the government should implement relevant policies that would inhibit the expansion of pollution-intensive industries in Argentina. As a result, economic globalization in the form of international trade participation would not significantly amplify the use of fossil fuel-generated electricity whereby the energy production-induced CO2 emissions can be controlled. Similarly, policies to restrict the inflows of dirty FDI can also reduce the demand for non-renewable electricity in Argentina; thus, the energy production-related CO2 emissions can be curbed further. Furthermore, it is more relevant for Argentina to boost its renewable electricity shares to reduce the energy production-based CO2 emissions associated with economic globalization.

Thirdly, the economic growth policies of Argentina should be aligned with its environmental development agendas so that the economic growth rate of the nation can be enhanced without boosting the energy production-related emissions. Once again, it is recommended that the bulk of the national output of Argentina is produced by combusting renewable electricity rather than employing the traditionally consumed fossil fuel-based electricity. In this regard, the Argentine government, rather than subsidizing the price of natural gas supplies, should subsidize the rate at which renewable energy is supplied to the national industries. Finally, the urbanization policies of Argentina should also be revisited to make the overall urbanization process environmentally sustainable. If the urbanization-induced energy demand in Argentina can be met using renewable electricity, then the energy production-related CO2 emissions associated with urbanization can be significantly reduced.

The major limitation endured in conducting this study was the unavailability of energy production-based CO2 emissions data beyond 2016. Besides, data unavailability also limited the choice of the control variables considered within the analysis. In the future, this study can be extended to assess the impacts of renewable electricity generation, economic globalization, economic growth, financial development, and urbanization on sector-specific CO2 emissions in the context of Argentina. The sectoral analysis would provide better insights for the policymakers to target certain sectors in which the fossil fuel dependency can be reduced to curb the associated CO2 emissions.