Sustainable pathways for attaining net-zero emissions in European emerging countries — the nexus between renewable energy sources and ecological footprint

This study aims to investigate the relationship between renewable energy and ecological footprint during the period of 1994–2018 from selected developing countries in Europe (Czechia, Croatia, Poland, Romania, Romania, and Turkey). In this context, the ecological footprint (EF), which has recently been the most widely used environmental indicator in the literature and is known as the most comprehensive because it includes many environmental factors, has been determined as the dependent variable. As independent variables, renewable energy consumption (REC), energy-related tax revenue (ETR), and energy productivity (EP) are included in the model. GDP and development of environment-related technologies (DET), which affect the ecological footprint in the model, are determined as control variables. As a result of the panel data analysis, according to the Durbin–Hausman cointegration test result, a long-term relationship between the variables was determined. According to the CCE estimator analysis, it can be said that there is a positive relationship between ETR and GDP variables and EF. For the AMG estimator analysis, it can be said that there is a positive relationship between GDP and EP variables and EF. Finally, according to the results of the Konya Causality test, a unidirectional causality relationship is detected from environmental technologies to the ecological footprint in Turkey, and a unidirectional causality relationship from the ecological footprint to GDP in Czechia, Romania, and Turkey. Furthermore, no causality relationship is detected between other variables. Based on the results, several policy implications are suggested.


Introduction
Recent research has generated a lot of discussion on the crucial issue of meeting climate change mitigation targets, which is also a crucial problem for decision-makers. Countries around the world set milestones for the transition to a post-carbon economy in terms of carbon neutrality, netzero carbon emissions, or even going above and above and considering being climate positive by reducing emissions more than they produce. There are many pressures that national economies must simultaneously address as they shift from carbon footprint to carbon handprint (Hand Print Action Towards Sustainability n.d.) at the national level, including enhancing social and economic features, ensuring energy security, and addressing inequality. Due to the necessity to address numerous development-related concerns at once, the policy mix is complicated.
According to the Paris Climate Agreement, there are only 29 years left to achieve global net-zero emissions. Depending on the state of the economy, different people are more or less inclined to contribute to this goal. Due to their complexity, social welfare, or energy efficiency, some of the world's economies are closer to the tipping point where they can easily reduce their environmental impact, while others continue to build their economic growth on the furnace-fuelled GDP. The final group, which consists mostly of rising economies, must drastically alter their plans for economic growth, adopt technological innovation in renewable energy sources, and sharply reduce their environmental impact. In order to minimize longterm emissions and meet climate change goals, the INDC (intended nationally determined contribution) mechanism was established after the Paris Agreement. Net zero policies outlined in the Paris Agreement at the national level play a significant role globally for reducing global warming to 1.5 °C (Runsen et al. 2022;Van Soest et al. 2021). The idea of net zero emission is to lessen society's influence on the environment and climate (Worth 2005;United Nations Environment Programme 2008). If global temperatures stabilize, the suggested targets should be attained (Intergovernmental Panel on Climate Change 2021. To meet the goal of 1.5 °C by 2050 and 2070, however, greenhouse gas (GHG) and CO 2 emissions must zero out (Climate Action Tracker 2021).
To reach net zero emissions by the goals of 2050 and 2070, more than merely energy-related emissions reductions are required. The developing nations of Europe have already made notable strides in this direction. Reducing material footprint and energy efficiency will make a big contribution to sustainable development. Due to low carbon resources and energy efficiency, development processes and CO 2 emissions have increased in response to ongoing energy demand (UNEP 2020). Although products and materials play a significant role in industrial emissions (Scott et al. 2018), However, the same point has been overlooked as a result of many climate change policies' emphases on energy efficiency and low-carbon energy technologies (Pauliuk et al. 2017). The inadequate plan for cyclical economics and material efficiency may hinder the effort to combat climate change (Pauliuk et al. 2017). As the paper focuses on some of the key tools, including renewable energy consumption (REC), energy-related tax revenue (ETR), energy productivity (EP), development of environmental technologies (DET), GDP growth, and ecological footprint (EF), it is crucial to investigate and work on various frameworks and instruments in this regard.
The energy crisis is placing tremendous pressure on the energy mix and could hasten the adoption of renewable energy sources. There is discussion about the social and economic viability of a genuine shift to renewable energy sources or diversification (Lampis et al. 2022). Renewable energy sources are essential for mitigating climate change and reducing carbon emissions, whether through expedited transition or broadened diversification. Additionally, there are nearly endless amounts of renewable energy sources. Fossil fuels are also subject to a volatile market that is influenced by a variety of circumstances, such as resource scarcity and armed conflicts like the one in Ukraine. After the collapse of the communist union, the economies of Eastern Europe suffered significant changes. They went from a centralised government to an expanded European Union as a result of these fundamental changes. These nations' regulatory systems were in disorder throughout the transition years, which resulted in uncontrolled economic development, increased resource depletion, and environmental mayhem. These nations, like other emerging and developing economies, prioritised economic growth over environmental protection (Danish and Khan 2020). When these Eastern European nations joined the European Union and had to abide by the environmental regulations set forth by the European Commission, the economic growth pattern was once again altered. Turkey, on the other hand, made additional strides and saw rapid economic expansion, still as a result of energy intensity and environmental deterioration. With the exception of Turkey, whose consumption has tripled over the past 30 years, all of the panel countries' energy consumption has remained at a level that is essentially stable or even slightly decreased. The move to renewable sources is the action to be taken to improve efficiency or reduce environmental damage caused by energy use in generally stable consumption settings. All of the chosen economies have made significant strides in this regard, with Poland and the Czech Republic raising their percentage of renewable sources by more than three times and Romania nearly tripling it. Turkey saw a more modest increase, from 24.9% in 2000 to 35.2% in 2021.
Our work is driven by the desire to use econometric analysis to analyses sustainable paths for achieving net zero emissions with the aid of renewable energy sources and ecological footprint in selected growing European nations, including Czechia, Croatia, Poland, Romania, and Turkey. In Europe, just 10% of recycled and reused materials are used (European Environment Agency 2019). As a result, there is still much work to be done to reach the zero-emission goal. The following ways that this research adds to the body of literature are the following: (i) There are few studies in the literature now on renewable energy sources and ecological footprint in relation to achieving zero emissions (e.g., Tiwari et al. 2023a;Shahzad et al. 2023;Tiwari et al. 2023b). Studies that concentrate on the relationship between renewable energy sources and ecological footprint disregard the problem of sustainable pathways. The sustainable routes, however, include sustainable development brought on by socio-cultural, environmental, and economic growth (Tiwari et al. 2022). In order to fill this gap, various research (Giljum et al. 2016;López et al. 2017;Greiff et al. 2017;Ma et al. 2018;Schandl et al. 2018;Jiang et al. 2019;Ansari et al. 2020;Karakaya et al. 2021;Wang et al. 2022) concentrated on ecological footprint and renewable resources. To the best of the authors' knowledge, this is the first attempt to achieve net zero emissions through a sustainable pathway that links ecological footprint (EF), GDP growth, development of environmental technologies (DET), and energy productivity (EP). (ii) For the policymakers of a few chosen countries within the context of the SDGs, the study suggests a number of methods for reducing emissions and achieving net zero emissions. Heterogeneity and cross-sectional dependence for certain nations and variables have been tested in advance of the empirical research.
The following is the study's plan. The introduction and theoretical underpinnings of the study are presented in the first section. The literature summary is included in the second section. The econometric techniques, methodology, and data estimations are introduced in the third section. The fourth section presents and examines the empirical findings, conclusions, and policy suggestions.

Literature review
The literature review section has been discussed under the following headings: (1) "Renewable energy vs. ecological footprint", (2) "Energy productivity and ecological footprint", (3) "Innovation and ecological footprint", and (4) "Green taxation", with the help of most recent literature.

Renewable energy vs. ecological footprint
The fundamental force behind reducing social inequalities is economic growth, and policymakers around the world are hesitant to sacrifice environmental quality in order to do so. However, as energy consumption boosts economic output, the switch to renewable energy sources supports both economic growth targets and environmental protection objectives, as previous research has shown that renewable energy has a smaller ecological footprint. Similar conclusions are discussed by Usman et al. (2021), who contend that using renewable energy slows down environmental deterioration while accelerating economic growth. Many developing nations maintain their economic growth by putting social and economic objectives ahead of environmental protection through their lax environmental regulation systems. In this perspective, the only choice for nations that must close structural and developmental gaps is to maintain high economic output while simultaneously reducing environmental damage. For such challenging goals to be met, increasing the use of renewable energy sources and greening the energy mix are essential. In order to stop environmental degradation and lessen their ecological impact, Sharma et al. (2021) present data in support of renewable energy promotion in nations with unsustainable economic development. In a recent study, Xu et al. (2022) show that renewable energy plays a crucial part in lowering environmental difficulties and is essential for doing so along with technological advancement. According to Huang et al. (2022), both emerging and developed nations continue to rely on conventional energy sources to produce their economic output, which is in conflict with environmental goals. Their ecological impact will be reduced as they switch to renewable energy sources, setting the stage for sustained economic growth.
Both developing and developed nations exhibit traditional economic growth patterns. The discrepancies are determined by the level of technological, social, or economic complexity. According to research by Ahmad et al. (2021), renewable energy reduces the environmental footprint but economic complexity increases it. The same study demonstrates how institutional integrity fosters sustainability by moderating the association between economic complexity and environmental footprint. The benefit for industrialised nations is that they have reached a stage of general development that enables progress to be reoriented or even slowed down. Developed economies, on the other hand, must close development disparities based on economic growth. Additionally, as energy consumption is a crucial component of the development of emerging countries, a growth model based on renewable energy sources is essential (Shahzad et al. 2021;Rosak-Szyrocka et al. 2023). Because sustained economic growth is mostly driven by energy consumption, efficient energy utilisation is essential. As many national economies struggle with low energy efficiency, Kızılgöl and Öndes (2022) believe that moving to renewable energy sources is essential for reducing ecological imprint. Access to conventional fuels may have an impact on the switch to renewable energy sources, while Adekoya et al. (2022) discovered that the relative quantity of fossil fuels makes the transition more difficult. At the same time, nations that import fossil fuels are more likely to transition to renewable sources since they are more energy efficient and leave a smaller environmental footprint.

Energy productivity and ecological footprint
Energy productivity was recently identified as one of the essential components for reducing environmental damage. Different energy use efficiency levels have an effect on carbon emissions, it was acknowledged. Increasing efficiency or higher energy consumption will lead to an expansion of economic activity. The promotion of policy alternatives in this area includes expanding the use of renewable energy sources, speeding up technological advancements in the energy sector, or implementing new green technologies (Xie and Jamaani 2022). Few prior research demonstrates a connection between ecological footprint and energy efficiency (Kazemzadeh et al. 2022). However, studies on the effects of energy productivity and efficiency on CO 2 emissions, environmental deterioration, and ecological footprint have been conducted by Kuittinen and Takano (2017), Ke et al. (2020), Özbuğday and Erbas (2015), Tajudeen et al. (2018), and Yao et al. (2021). Among these authors, Tajudeen et al. (2018) and Yao et al. (2021) looked into how energy production and efficiency reduce ecological footprint and environmental deterioration. Yao et al. (2021) also provided evidence of the relationship between corruption, energy efficiency, and ecological footprint for the BRICS countries between 1995 and 2014. They made use of data envelopment analysis (DEA) models and the generalised method of moments (GMM). They discovered that corruption makes it more probable for the ecological footprint to shrink and for energy efficiency to rise. However, with the help of green technical advancements and resource rent responses, both environmental quality and energy efficiency could be enhanced simultaneously. The causality findings that link corruption control, industrialization, ecological footprint, trade, energy efficiency, green technology innovation, natural resource rent response, and financial development provide support for the feedback hypothesis. Tajudeen et al. (2018) looked at the increases in energy efficiency brought on by environmental deterioration and CO 2 emissions for the 30 OECD (Organisation for Economic Cooperation and Development) nations. According to research, energy efficiency is vital for reducing environmental degradation and CO 2 emissions, and money has a significant positive impact on both, i.e., environmental degradation and CO 2 emissions. The majority of these studies have not looked into any energy productivity-related topics. As a result, we chose energy production as one of the crucial factors in our research. However, as discovered by a different team of researchers, energy efficiency causes a rise in CO 2 emissions, environmental deterioration, and ecological footprint (e.g., Ke et al. 2020;Kuittinen and Takano 2017;Özbuğday and Erbas 2015). Kuittinen and Takano (2017) investigated the relationship between efficiency and carbon footprint in 2011 in temporary housing in Japan. In light of energy life cycle assessment and simulation, the authors proposed three alternative shelter models: sea container shelters, prefabricated shelters, and hardwood log shelters. Due to their significant life cycle emission contributions, the aforementioned building materials are best suited for usage as temporary housing. Ke et al. (2020) investigated the effect of energy efficiency on the ecological footprint in 280 Chinese cities between 2014 and 2018. To investigate changes in ecological footprint brought on by innovation effectiveness at various economic development levels, they employed threshold regression and generalised spatial two-stage least squares (GS2SLS) models. For the Central and Eastern regions of particular Chinese cities, it was shown that there was a considerable inverse link between ecological footprint and energy efficiency. That example, improvements in energy efficiency reduce the ecological impact.

Innovation and ecological footprint
Emerging economies must continue to increase their economic output while also finding ways to green this output and reduce their ecological imprint. Technology improvement has raised productivity and produced more affordable items under the conventional growth paradigm. The necessity to increase economic sustainability in terms of environmental preservation, however, claims additional significance for encouraging innovation. The greatest policy options for emerging economies are those that will both reduce their environmental impact and generate the required economic value for them to close the gap with industrialised economies. According to Ahmad et al. (2020), technological advancement and innovation are some of the finest ways to do this. Renewable energy sources, as previously said (Abban et al. 2022), are essential for cutting carbon emissions. However, the transition thereto continues to be a focus of significant financial efforts for creating such technologies and revolutionising the sector. To generate a greater proportion of renewable energy sources while reducing the ecological impact, however, the financial efforts are worthwhile (Altintaș and Kassouri 2020). As Suki et al. (2022) show, combining technology advancement with renewable energy sources is an effective environmental protection policy tool for a seamless green transition because it results in a decrease in carbon emissions and ecological footprint.
Regarding an ecological footprint and technological advancements, a strong correlation was found. Authors from several disciplines looked into both positive and negative associations. Green energy minimizes energy consumption by substituting fossil fuels, while green technological innovation (GTI) advances environmentally and humanely friendly goods and energy. Additionally, GTI greatly lowers CO 2 emissions and contributes to a better environment (Shahzad et al. 2020). Technology advancements have enhanced and made garbage disposal processes more environmentally friendly in various situations (Tercan et al. 2015;Tsai and Kuo 2010). Municipal solid waste (MSW) management and landfill disposal procedures are more feasibly achieved through value-added strategies using contemporary technologies. Because MSW has a high organic content and moderate energy level, it can be used to generate electricity by technological means (Consonni et al. 2005;Pan et al. 2023;Apostu et al. 2022). While negatively, industrialization, mass production, and technical advancement are severely stressing the biosphere and the natural world (Galli et al. 2012). Because of our over-reliance on technology and contemporary lifestyles, the economy-environment dichotomy is simple to identify. In other words, there is a large gap between the demand for resources, which is limitless, and the supply, which is also restricted. Furthermore, the development of green technologies and the excessive use of technical tools and equipment are to some extent exerting pressure on the ecological environment. Therefore, it is crucial to strike a balance between technical advancements and ecological well-being through the efficient use of resources.

Green taxation
In many circumstances, one public policy instrument for resolving ecological challenges is environmental taxation, which is a manner of internalising harmful environmental externalities. Taxes and subsidies from the government are examples of environmental intervention. Both components are essential for encouraging a specific behaviour or putting into practise green public policies. Government financial intervention may be used to promote energy efficiency or the switch to renewable energy sources. It takes a lot of effort to advance new trends, like renewable energy, especially when financial efforts are anticipated. The move to a greener economy is driven by new technology and innovation and requires significant financial investment. Governments can encourage it in two ways: by providing incentives to those who commit to the process or by levying taxes on those who do not perform well for the environment and do not make efforts to improve it. By doing so, the governments will use the money they have gained by taxing unclean and wasteful energy practises to fund energy technical innovation efforts. However, authorities must pay close attention to the energy sector's advancements and the objectives of environmental preservation, and the tax system must be updated frequently. Without it, the fiscal structure is unable to support the switch to renewable energy sources and the reduction of carbon emissions (Dogan et al. 2023). This strategy could have a problem because it might be less expensive to pay for the carbon one produces than to invest in a greener route. Additionally, the revenues obtained through environmental taxation are mostly driven by changes in economic production rather than by harsher regulations or taxation rates (Andreoni 2019), which leads us to conclude that environmental policies can still be improved. One method to resolve this issue is to separate pollution costs from environmental protection levies. According to Fang et al. (2023), this distinction lowers the ecological footprint whereas environmental protection levies raise it. The fairness of carbon taxes is another topic of discussion because they could exacerbate social inequality and cause greater problems with inclusivity (Bourgeois et al. 2021;Sommer et al. 2022). In addition, raising carbon taxes could slow economic expansion and raise poverty levels (Mardones and Mena 2020). However, as Dogan et al. (2022) show in a recent analysis, environmental taxation has a moderating effect on carbon emissions in the case of industrialised nations, giving green technology a competitive edge and raising taxes on the operational expenses of dirty ones. Having stated that, environmental taxes must strike a balance between viewpoints on economic growth and alleviate ecological degradation issues while also funding green innovation subsidies and environmentally conscious behaviour (Liu and Ge 2023).

Econometric methods and methodology
The long-term relationship between renewable energy and ecological footprint is examined in the study's analysis for five selected developing nations in Europe (Czechia, Croatia, Poland, Romania, Romania, and Turkey). The tests are conducted utilising new-generation panel data analysis methods and annual data from 1994 to 2018. The central claim of the study is that "renewable energy sources and ecological footprint have a long-term relationship". In this context, the approach to be utilised is decided upon once the data set and model of the variables to be employed in light of the 1 3 hypothesis are introduced. The results of the analyses are evaluated once the theoretical and conceptual framework for the tests to be used within the parameters of the procedure is presented.

The data set and model
Due to a shared data restriction across the variables in the model, the analyses are based on annual data for the years 1994 to 2018. As a result, the five developing European nations of Croatia, Poland, Romania, Romania, and Turkey are chosen as the study's country sample. This group of nations was chosen mostly due to their recent rise to prominence in both European and international trade, strong potential for expansion, and status as the nation's receiving the greatest foreign investment following the pandemic. These nations also stand out in the environmental issues brought on by global industry, given their population concentrations and market sizes. The kind and quantity of energy utilised during production phases, as well as the environmental rules they will put in place, are crucial in today's interconnected world. As a result, the sample of the study is made up of these five nations, who are anticipated to lead the way in future global trade.
Based on studies in the literature, the variables in the model developed to test the hypothesis were chosen. The ecological footprint (EF), which has lately been the most frequently cited environmental indicator in the literature and is regarded as the most thorough because it takes a wide range of environmental issues into account, has been chosen as the dependent variable in this situation. The model includes three independent variables: energy productivity (EP), tax revenue connected to energy (ETR), and consumption of renewable energy (REC). The ecological footprint in the model is affected by the GDP and the development of environment-related technologies (DET), which are chosen as control variables.
All of the model's variables are the most desired variables and do not have widespread data issues, as can be shown from the literature. The most commonly chosen control variable in environmental research, particularly in EKC hypothesis analysis, when these variables are examined, as observed in many studies in the literature, is GDP. Since it is believed that advancements in environmental technologies will have an impact on the ecological footprint, DET, which is chosen as another control variable, is incorporated into the model. In addition, it is not essential to take the variables' logarithms because every variable in the model is composed of proportional expressions. Table 1 lists relevant variables and any necessary descriptions.
In the paper examining the relationship between renewable energy and ecological footprint, the model created within the specified sample and data range is constructed as follows within the scope of the hypothesis established.
In the model, i = 1, 2, 3, ....N denotes cross-section data, t = 1, 2, 3, .....T denotes time dimension, and ɛ denotes the error term. Since all variables in the model are in the form of proportions or indices, they are included in the analyses without taking their logarithms.

Econometric method
The following methodological steps were used in the study to examine the long-term relationship between renewable energy and ecological footprint over five selected developing countries in Europe: Analysis of graphs and descriptive statistics of variables; Breusch-Pagan's CDlm1 and Pesaran et al.'s LMadj test for analysing the existence of cross-section dependence of variables; and LMadj test for analysing the existence of the relationship between variables and ecological footprint. Fisher ADF, Fisher PP, Im et al. (2003) unit root tests; the (PANIC) unit root test developed by Ng (2004, 2010), Levin et al. (2002), and Levin et al. (2002), applying the Durbin-Hausman cointegration test created by Westerlund (2008) to assess whether there is a cointegration relationship between variables; the Delta test established by  to determine whether the slope coefficients vary between units; the CCE estimator (1)  Pesaran (2006); and the AMG estimator developed by Eberhardt and Bond (2009) are used to estimate the cointegration coefficients of the variables. Lastly, the Kónya (2006) panel causality test is used to determine whether the relationship between the variables is causal.

Descriptive statistics and graphical analysis of variables
Before moving on to the analyses in econometric studies, the descriptive statistics for each of the variables in the model should be provided and interpreted independently. In this setting, it is possible to see the changes and cyclical oscillations of the variables between the given years and comment on their statistical changes. is interpreted according to the kurtosis and skewness results. In cases where the kurtosis value is greater than 3, it can be said that the series is pointed, and in cases where it is less than 3, it can be said that the series is flattened. Considering the interpretation of skewness result, when the value is equal to zero, it shows normal distribution. A value greater than zero indicates that the series is positively (left) skewed, while a value less than zero indicates that the series is negatively (right) skewed (Kapusuzoglu and Karan 2010: 61-62).
According to the results of Table 2, since the skewness values of EF, REC, and DET variables are greater than zero, the series are left skewed, while the values of other variables are right skewed since they are less than zero. In kurtosis values, since the values of DET and GDP variables are greater than 3, the series are pointed, and since the values of other variables are less than 3, the series are flattened.

Cross-section dependency test
Prior to conducting hypothesis testing in research using panel data analysis, it is required to ascertain whether there is a cross-sectional relationship between the variables. The interdependence of nations grows as the globe becomes more globalised every day. As a result of this interdependent mechanism, positive or negative shocks in one country may have an impact on another. Due to the common factor problem, it is important to know the cross-sectional dependency of variables in econometric investigations When the time dimension is larger than the cross-section dimension (T > N), the Breusch and Pagan (1980) CDlm1 test is used to detect cross-section dependence. When the time dimension is equal to the crosssection dimension (T = N), the Pesaran (2004) CDlm2 test is used to detect cross-section dependence. Finally, when the time dimension is both smaller (TN) and larger (T > N) than the cross-section dimension, the Pesaran (2004) C Five countries make up the nation group analysed in the research. Consequently, N = 5 is the cross-sectional dimension. Since the years between 1994 and 2018 are being examined, the time dimension is 25 (T = 29). As a result, the observation dimension is smaller than the temporal dimension. Both the CDlm1 test by Breusch and Pagan (1980) and the LMadj test by  are employed in the analyses since T > N. When Table 3, which shows the cross-section dependence test results, is analysed, it is seen that the probability values of all variables except the DET variable are statistically significant at the 1% level. Accordingly, for all variables, the null hypothesis "there is no cross-sectional dependence" is rejected and the hypothesis "there is cross-sectional dependence among countries" in the panel data is accepted. This situation is also compatible with today's global world, and it is concluded that a shock to one of the selected 5 country groups will also affect other countries. For this reason, the authorities and decision makers of the countries in the sample included in the analysis should steer the future by taking into account the current situation. The probability value of the DET variable is statistically insignificant according to the LM adj test results. In other words, there is no crosssection dependence for the DET variable.

Panel unit root test results
In econometric analyses, stationarity tests are required to solve the problem of spurious regression. Granger and Newbold (1974) state that if there is a unit root in the series of the variables included in the model, the results obtained from the analyses will not be actual findings. The main issue to be considered in the stationarity tests of panel data analyses is whether the countries in the sample included in the model are independent of each other. In this context, unit root tests of panel data analyses consist of first-and second-generation tests. While the first generation of unit root tests do not take cross-section dependence into account, the second-generation tests perform their analyses according to cross-section dependence. In today's global world, it is more realistic that a shock to one of the countries that make up the panel will also affect other countries, and the use of second-generation tests used in the literature is therefore interpreted as a more realistic approach. Since cross-sectional dependence is observed among the variables included in the model, PANIC unit root test, one of the second-generation unit root tests, is used for all variables. This test, developed by Ng (2004, 2010), analyses the stationarity of residuals and factors separately. This test is known in the literature as panel analysis of stationarity of residuals and common factors (PANIC). The data generating process is as follows for variable X: The variable X it is the sum of the common factor and the residuals. The variable F t is used to eliminate the crosssectional dependence problem. Factor estimates are obtained by applying the principal components method to the first differenced data. Consistent estimation of factors regardless of whether the residuals are stationary or not does not require (2) X it = D it + � i F t + e it the residuals to be stationary. The advantage of this test is that it tests for the presence of a unit root in the residuals when the unit root in the factors is rejected.
For the stationarity of the residuals, P a and P b PANIC test statistics are used. It is constructed from the p values of the ADF test statistics that investigate the stationarity of e it individually. P a shows the results of the ADF test with constant, and P b shows the results of the ADF test with constant and trend. In addition, Panel-adjusted Sargan and Bhargava (PMSB) test used for the autocorrelated e it case. If any of the P a , P b , and PMSB statistics are unit rooted, it is concluded that the variable is unit rooted.
In PANIC unit root test results, all three probability values Pa, Pb, and PSMB should be statistically significant. If one of the statistical values is not significant, it is interpreted that the variable has a unit root. In this context, it is seen that all variables included in the model have unit roots in both constant and constant-trend models, and the variables become stationary when the difference is taken (Table 4).
Since there is no cross-section dependence in the DET variable included in the model, its stationarity is analysed with first-generation unit root tests (Im, Pesaran and Shin, Levin, Lin and Chu, Fisher ADF, and Fisher PP). In these tests, a probability value close to 0 means that the series are stationary, while a probability value close to 1 means that there is a unit root in the series. Table 5 shows the unit root test results for DET variable in constant and constant-trend models. According to the results, DET variable is stationary at 1% significance level in all tests.
The stationarity analysis results in Table 5 can be interpreted as the shock to one of the countries included in the model creates permanent results and does not lose its effect immediately. Moreover, non-stationarity of the series provides the necessary precondition for cointegration tests. When the same test is repeated by taking the first-order difference of all series for the stationarity of the series, it is concluded that the variables become stationary at I(I) level. DET variable is stationary at I(0) level.
In the Durbin-Hausman cointegration test, which is the cointegration test to be used in the next section of the study, the dependent variable being I(I) indicates that the sufficient condition for the analysis is met.

Homogeneity test
In panel data analysis methods, it has to be decided whether the coefficients of the variables assumed to have a long-run cointegration relationship are homogeneous or not. The homogeneity test tests whether the change in one of the countries affects the other countries at the same level. In this context, coefficients are expected to be heterogeneous in models constructed for countries with different economic structures, whereas coefficients are expected to be homogeneous in models constructed for country groups with similar economic structures. In this paper, the Slope Homogeneity Test (Delta test) developed by  is used to test homogeneity. The Delta test is valid for large samples, and the Delta adj. test is valid for small samples. In the homogeneity test, the null hypothesis (H 0 ) is interpreted as "slope coefficients are homogeneous" and the alternative hypothesis (H 1 ) is interpreted as "slope coefficients are heterogeneous".
The homogeneity test results of the variables included in the model in the study analysing the relationship between renewable energy and ecological footprint are given in Table 6.
According to the homogeneity test results in Table 6, the H 0 hypothesis based on the homogeneity of the coefficients in the Delta test is rejected at 1% significance level, and it is decided that the coefficients are heterogeneous. This situation reveals that the effect of a change in the variables included in the model on the ecological footprint differs from country to country.

Durbin-Hausman cointegration test results
The cointegration connection should be checked for the presence of a long-run relationship after the stationarity of the variables has been established. The approaches most commonly employed in the literature-namely, those of Pedroni (1999), Pedroni (2007), Westerlund (2008), and Westerlund and Edgerton (2007)-are used to examine whether a long-run association exists in panel data analysis. Cross-section dependence must be included in cointegration analyses, just as it is in unit root tests. Otherwise, issues could arise, such as embracing the idea that there is a cointegration link even while there is none. This issue led Westerlund (2008) to develop the Durbin-Hausman analysis, which is used in this study and takes into account cross-section dependence. The Durbin-Hausman (DH) test created by Westerlund (2008) was chosen for this investigation for a variety of reasons. The test's second-generation panel cointegration design, which takes into consideration cross-sectional dependence, is by far its most significant advantage. The dependent variable must be I(I), but the independent variables may be I(0) or I(I) (Westerlund 2008: 205). In addition to this, the Durbin-Hausman cointegration test permits the panel's parameters to be both the same (homogeneous) across units and different (heterogeneous). If the parameters are homogeneous across units, the DH panel test statistic is utilised, and if they are heterogeneous, the DH group test statistic is used. According to the findings of the Delta test devised by , it is concluded that the coefficients are diverse in the study examining the relationship between renewable energy and ecological footprint. Therefore, it can be said that the findings of the DH group test statistic give more trustworthy results for the cointegration test. The cointegration connection can be examined separately in the panel dimension and in the group dimension using the Durbin-Hausman approach. The autoregressive parameter may vary across cross sections in the DH group test. A cointegration connection may occur in some cross sections if the H_0 hypothesis is rejected, according to this test. This test makes the assumption that the autoregressive parameter is constant throughout all cross sections. This presumption states that the cointegration relationship is considered to exist for all cross sections when the H_0 hypothesis is rejected (Di Iorio and Fachin 2007). Within the parameters of this test, the relationship between renewable energy and ecological footprint is examined. The findings are presented in Table 7.
Since it is determined that it would be more appropriate to use group statistics in the study according to the result that the slope coefficients change and the variables are heterogeneous in Table 7, the results of Durbin-H Group statistics are taken into consideration. When the probability values of the Durbin-H Panel statistic are analysed, it is concluded that there is a long-term relationship between the variables since it is less than 0.05. Therefore, it is concluded that there is a long-term relationship between renewable energy and ecological footprint in the model.
The detection of long-run relationships between the variables indicates that the necessary precondition for coefficient estimation is met. After the cointegration relationship between two variables is detected, Common Correlated Effects estimators developed by Pesaran (2006) and AMG estimators developed by Eberhardt and Teal (2010) are used to estimate cointegration coefficients.
The Monte Carlo study by Pesaran (2006) shows that crosssection dependence should be tested in panel data models and methods that take this into account, if any, should be used. The Common Correlated Effects (CCE) estimator is an estimator that takes into account the dependence between the cross sections forming the panel and is developed by Pesaran (2006). CCE longrun coefficient estimators assume that the independent variables and unobserved common effects are stationary and exogenous. It is also consistent when the independent variables and unobserved common effects are stationary (I(0)), first-order integrated (I(1)), and/or cointegrated. Eberhardt and Bond (2009), Eberhardt and Teal (2010), developed the Augmented Mean Group (AMG) estimator, which considers cross-sectional dependence. The AMG estimator takes into account the differences in observable and unobservable factors between panel groups as well as time series characteristics. Eberhardt and Bond (2009) developed an estimator that can calculate cointegration coefficients for the countries forming the panel and the overall panel with the AMG test. In this method, it takes into account the common factors in the series and is also used in the presence of endogeneity problem, which indicates that there is a correlation between explanatory variables and error terms (Eberhardt and Bond 2009). AMG estimators with cross-sectional group specification are calculated by averaging the coefficients of each country in the panel. This test is also more powerful than other coefficient estimation methods as it estimates the arithmetic mean of the cointegration coefficients by weighting. The results of AMG and CCE tests are shown in Table 8. According to the CCE estimator analysis, the coefficient of the ETR variable is significant at the 5% level and the coefficient of the GDP variable is significant at the 10% level. Based on this result, it can be said that there is a positive relationship between ETR and GDP variables and EF. This result is consistent with the theory. Because many studies in the literature have proved that there is a positive relationship between GDP and EF within the scope of the EKC hypothesis. The same is true between tax revenue in the energy sector and EF. Therefore, the positive coefficients of the two variables are within the expectations. The coefficients of other variables are statistically insignificant.
According to the AMG estimator analysis, the coefficient of the GDP variable is significant at the 1% level and the coefficient of the EP variable is significant at the 10% level. Based on this result, it can be said that there is a positive relationship between GDP and EP variables and EF. This result is also consistent with the theory. An explanation about GDP has been made above. Similar processes are also valid for EP. It is known that improvement in energy efficiency will also have a positive effect on EF. Therefore, the positive coefficients of these two variables are within expectations. The coefficients of other variables are statistically insignificant. Kónya (2006) developed this test that investigates the existence of causal relationships between variables by using the Seemingly Unrelated Regressions (SUR) estimator introduced to the literature by Zellner (1962). One of the advantages of this test is that since the panel is assumed to be heterogeneous, causality tests can be applied separately for the countries belonging to the panel. Another important advantage of this test is that there is no need to apply unit root and cointegration tests since country-specific critical values are generated. Suppose the Wald statistic calculated for each country after applying the previous test is greater than the critical values at the significance level. In that case, the null hypothesis, "there is no causality between the variables", is rejected. In other words, when the Wald statistic is greater than the critical value, it is concluded that there is causality between the variables.

Kónya causality test
According to the causality analysis results between energy tax revenue and ecological footprint in Table 9, no causality relationship is detected between the variables in any country.
In Table 10, only in Turkey, a unidirectional causality relationship is detected from environmental technologies to ecological footprint at 10% level. No causality relationship is detected between other variables. Based on this result, it can be concluded that the technology-intensive production structure in Turkey may be effective on ecological footprint.  Table 11 shows a unidirectional causality relationship from ecological footprint to GDP at 10% level in Czechia and Romania and at 5% level in Turkey. It is concluded that ecological footprint is effective with the production dimension in the mentioned countries, and that the reduction of environmental problems can be effective in production and national income by creating an exogenous effect.
According to the causality analysis results between energy efficiency and ecological footprint in Table 12, no causality relationship is detected between the variables in any country.

Conclusion and policy implications
This study aims to measure the relationship between ecological footprint (EF) and renewable energy (RE) from selected developing nations in Europe (Czechia, Croatia, Poland, Romania, Romania, and Turkey) between the years 1994 and 2018. The most often used environmental indicator in the literature at this time is EF, which is determined as a dependent variable and comprises a variety of environmental parameters. The following proxy variables, i.e., renewable energy consumption (REC), energyrelated tax revenue (ETR), and energy productivity (EP), are calculated under the independent variable, i.e., renewable energy (RE). Since both GDP and development and environment-related technologies (DET) have an impact on the ecological footprint of the model, they are both regarded as control variables. The Durbin-Hausman cointegration test was used in panel data analysis to identify a long-term link between the variables. ETR, GDP, development, and ecological footprint all appeared to be positively correlated, according to the CCE estimator analysis. Similar to this, an analysis using the AMG estimator found a link between energy productivity (EP), GDP, and ecological footprint. The Konya causality test also reveals a unidirectional causality link between environmental technologies and Turkey's ecological impact. An estimated unidirectional causality association between ecological footprint and GDP has been found for the countries of Turkey, Romania, and the Czech Republic. Furthermore, no causal connection between any other study factors has been found. Therefore, governments and associated stakeholders of certain emerging European nations must implement various types of interventions in line with the policy perspective in order to reduce the ecological footprint and attain net zero emissions. Humans' use of commodities and services to meet their needs leads to significant environmental degradation and destruction, or EF. Humans typically have a variety of additional options at their disposal to meet their demands. Therefore, it is crucial for politicians to educate the public on how to effectively use limited resources, the sustainability of their lives, and how to alter their patterns of consumption of goods and services. In order to reduce CO 2 emissions, improve ecological sustainability, and enlist the participation of individuals, numerous legislative measures are needed. The first step is to keep the production and consumption of products and services in balance. According to our research, the GDP, EF, and energy-related tax revenue (ETR) are all positively correlated. In order to reduce ecological footprint without compromising economic development, governments of the various nations should concentrate on reducing GDP and tax revenue tied to energy. In addition, further efforts are required to limit human fertility while keeping in mind the planet's limited ecological capacity. The rise in ecological footprint is positively caused by energy output as well as GDP, as seen in the data. Therefore, it is imperative to investigate advancements in green technology, green finance, and renewable energy sources. A unidirectional causality association between environmental technologies and ecological footprint is found in the instance of Turkey. That demonstrates the urgent need for responsible authorities to support green technologies and energy sources. For Czechia, Romania, and Turkey, a unidirectional causality relationship between ecological footprint and GDP is estimated. Therefore, utilising resources (labour, capital, and energy) efficiently can lower ecological footprint without harming economic development and growth. Furthermore, regulations that support green technologies, efficient energy use, and renewable energy help to reduce the ecological footprint. Finally, trade openness promotes economic prosperity and lessens environmental impact to achieve net zero emissions.

Data Availability
On request data is available with corresponding author.

Declarations
Ethics approval Ethical approval is not applicable as the data is obtained from different databases, and no questionnaires for animals or humans are used.