The moderating role of technological innovation and renewable energy on CO2 emission in O.E.C.D. countries: evidence from panel quantile regression approach

Abstract This study analysed data from a panel consisting of 32 O.E.C.D. member countries for the years 1996–2020. This research explores the nexus between CO2 emissions, G.D.P. per capita, renewable energy supply, the development of patents, and gross fixed capital formation in the context of 32 O.E.C.D. countries. Also, the panel quantile regression technique is being used to investigate potential variations in heterogeneity and asymmetry. The empirical evidence shows that technological innovation negatively impacts CO2 emissions; however, the impact varies greatly between quantiles. This research also explores the potential for heterogeneity and asymmetry in the moderating effect of technological innovation with regards to economic growth and renewable energy. The investigation, which relied on the use of panel quantile regression, revealed that technological innovation exerts a wide variety of moderating effects. In conclusion, the study provided policy recommendations.


Introduction
The 2015 United Nations Climate Change Conference (COP21) in Paris emphasised the importance of reviving public and private activities to boost the creation of greener energy sources.The O.E.C.D. countries that have signed on to this initiative have committed to further strengthening their collaboration on research and innovation initiatives with the goal of enhancing climate patterns.Since they have the largest economies and energy consumption, O.E.C.D. countries contribute disproportionately more to CONTACT Najia Saqib nsaqib@psu.edu.saglobal warming than any other group of countries.The countries in the O.E.C.D. were responsible for about 32.50% of the world's total carbon emissions, which added up to about 10,292.73Mt. Figure 1 shows that renewable energy accounts for 10.95% of the world's primary energy supply and also environmental technology patents are increasing at a rate of 10.43% annually (OECD, 2020).
A synergistic plan to manage excessive levels of CO 2 emissions is required in order to tackle challenges related to global warming and other vulnerabilities to the environment.It is possible that investments in innovation and technology will prove to be an efficient approach.This is due to the fact that the development of environmentally friendly innovations and technologies is necessary for the reduction of carbon emissions and the promotion of the growth of eco-friendly economic systems (Ganda, 2019;Ulucak et al., 2020).With the goal of keeping global warming well below 2 C, the Paris Climate Change Conference (COP21) and the Paris Agreement (UNFCCC, 2015) emphasise the need to eliminate CO 2 emissions.It is evident that environmentally friendly innovations are needed to reach this goal, and activities encouraging clean energy technology through innovation processes are essential.Security of energy supply and climate change mitigation both depend on such measures, as both energy consumption and carbon emissions have been increasing rise in recent years (Erdo gan et al., 2021;Wang et al., 2012).Increases in environmental damage have been met with a variety of responses during the past few decades, including the creation of new technologies and the filing of patents linked to such technologies.Entrepreneurs may not be motivated by environmental technology innovations that are crucial for the mitigation of climate change (Zhang et al., 2016).Hence, strict and robust legislative measures are essential to regulate the large costs associated with creating renewable technologies (Su & Moaniba, 2017).Environmental technology innovation will be available primarily to countries with high per capita incomes (Du et al., 2019), though countries with the right policies in place may be able to bridge this gap regardless of wealth (Koc ¸ak et al., 2019).
Investments in renewable energy sources are of essential importance for achieving economic, social, and environmental sustainability (Ulucak et al., 2019) because of the central role that energy plays in sustainable development.As a result, green energy has been emphasised as having a crucial role in fostering economic growth, decreasing pollution and bettering society (Ulucak et al., 2019).Consumption of renewable energy, which incorporates sustainable technologies, is preferable to the use of fossil fuels, which are the primary source of greenhouse gas emissions (Ganda, 2019).All parts of the economy need cleaner technologies to increase the use of renewable energy and protect energy sources (Mensah et al., 2019).
More research and inventions, which play a significant role in reducing emissions, are needed in light of their dramatic ascent in recent years.Since patent counts are considered one of the most essential indicators of innovations and also a measurement for technical advancement (Lindman & S€ oderholm, 2016), studies in this context have focused on the patent's role as a representative for sustainable technological innovations (Du et al., 2019;Mensah et al., 2019).Different viewpoints have been presented in the literature on the topic of technological innovation and the CO 2 emission nexus, but the importance of technological innovation in reducing carbon emissions remains debatable.Some studies back up what was thought to be the expected effect of innovations (Su & Moaniba, 2017;Wang et al., 2012Wang et al., , 2019)), while others find that innovations have a small effect on emissions or help raise them (Du et al., 2019;Hodson et al., 2018;Mensah et al., 2018).
It is crucial to learn more about the chain of events that begins with technological innovation and concludes with economic performance and carbon emissions by using cutting-edge research methods.This study analyses the effects of technological innovations on CO 2 emissions in O.E.C.D. countries, taking into account the role of renewable energy, G.D.P., and investment within a framework developed specifically for this purpose.This study investigated at how new technologies have changed CO 2 emissions in 32 O.E.C.D. countries between 1991 and 2020.This research uses the panel quantile regression method in order to further investigate the dilemma of whether or not the influence of technological innovation on CO 2 emissions in O.E.C.D. countries is heterogeneous and asymmetric.This is done in light of the fact that there are significant differences between the various countries that make up the O.E.C.D..The following are the primary contributions that can be drawn from this research: (1) This research provides fresh evidence for designing efficient strategies to reduce emissions and encourage technological advancement in O.E.C.D. member countries, as well as a more in-depth analysis of the effect of technological innovation on CO 2 emissions in these countries; (2) In the majority of prior studies, researchers have neglected to take into account the likelihood of heterogeneity and asymmetry in the relationship between technological innovation and CO 2 emissions in O.E.C.D. member countries, leading to skewed and contradictory results (Gozgor et al., 2018).We use the panel quantile regression method to look into the possibility of heterogeneity and asymmetry to address this limitation.
The remaining sections of the study are as follows: In the Section 2, we analyse the literature that has come before.Methodology, including model design, estimating strategy and data, are presented in Section 3. The Section 4 contains the discussion and reporting of the empirical findings.Several policy implications are discussed in the final section of this article.

Literature review
As the intensity of technological innovation persists, more and more government officials and academics recognise the significance of technological innovation for reducing CO 2 emissions (Chen et al., 2018;Saqib, 2022c;Sharif et al., 2022;Song et al., 2019).For this reason, many academics have begun studying how technological innovation affects carbon dioxide (CO 2 ) emission.The growth of patents is one widely-used indicator of technical innovation that has sprung from this research.To begin, some academics have attempted to quantify the level of technological innovation by looking at energy savings and R&D spending, and they have also investigated how these factors affect carbon dioxide emissions.Firstly, the impact of patents on carbon dioxide emissions has been extensively studied since patent growth is typically used as a proxy for technological innovation, such as the works of Alvarez-Herr anz et al. (2017), Dong et al. (2020), Hashmi and Alam (2019), Wang et al. (2019) and Wurlod and Noailly (2018) also came to identical conclusions on the relationship between energy efficiency and CO 2 emissions.Some researchers explore the dynamic links connecting research & development and CO 2 emissions, as the research & development scale is also considered an important indicator for gauging technological innovation trends (Churchill et al., 2018;Fern andez Fern andez et al., 2018;Petrovi c & Lobanov, 2020).The findings of these research are extremely important since they reveal the effects of technological progress on CO 2 emissions.Despite this, study did not present a comprehensive analysis of the mechanism of influence that technological innovation has on carbon emissions.
An expert contends that further pollution can be reduced without a corresponding reduction in productivity if the amount of energy used in industrial processes is controlled (Requate & Unold, 2003;Zhang et al., 2022).The development of technologies that are less harmful to the environment can also contribute to increasing the efficiency with which energy is generated (Ahmad et al., 2022;Goulder & Mathai, 2000;Lantz & Feng, 2006;Pao & Tsai, 2011;Peng et al., 2022;Saqib, Duran, et al., 2022;Socolow et al., 2004;Tang et al., 2022).Climate policy can promote the economy in two distinct ways: first, by making use of or investing in already existing 'eco-friendly' technologies; and second, by funding R&D efforts so, environmental laws that tax businesses that pollute and give subsidies to businesses that use technology that is good for the environment can encourage and help the development of these technologies (Irfan et al., 2022;Jin et al., 2022;Markewitz et al., 2012;Li et al., 2022;Saqib, Sharif, et al., 2022).The CO 2 emission can be reduced with environmentally sustainable technologies, according to numerous studies, which in turn promotes sustainable economic growth (Kaygusuz et al., 2007;Saqib, 2022a;Yin et al., 2015) and reach the UN's SDG-7 and SDG-13 targets.
Since Grossman and Krueger (1991) established the environmental Kuznets curve (E.K.C.) hypothesis to evaluate the connection between economic growth and environmental pollution levels, many researchers have started to estimate the effect of economic growth on CO 2 emissions using the E.K.C. hypothesis (Alam et al., 2019;Munir et al., 2020).Numerous more previous research accept the inverted U-shaped E.K.C. concept, including Atici (2009) for Central and Eastern Europe, Al-Mulali et al. (2015) and Al-Mulali and Ozturk (2016) for advanced countries and Rauf et al. (2018) for Belt and Road Initiative countries.Contrary to this, some past studies contradict the E.K.C. concept, including Ozcan (2013) for Middle East countries.
Second, the renewable energy supply sources are often cited as a key component that can decrease CO 2 emissions; this viewpoint has been confirmed by Li et al. (2020) and Wang, Dong, et al. (2020).Most researchers agree that renewable energy sources are among the most sustainable when it comes to the health of the environment and the economy (Saqib, 2018;Yang et al., 2022).Evidenced by Sarkodie and Strezov (2019) as well as studies like Ali et al. (2019) and € Ozokcu and € Ozdemir (2017) which discovered that overall pollution level in O.E.C.D. countries increased dramatically between 1980 and 2010 due to higher energy usage; in Nigeria.According to the research conducted by Ali et al. (2019) the use of energy has a good effect on the environment, whereas replenishing energy sources has the adverse impact.Based on concrete fact, it is clear that using renewable energy sources can significantly boost environmental quality over time.Renewable energy, according to previous research by Saqib (2022c) results in a considerable decrease in pollution levels and carbon emissions, both of which have a positive effect on the environment.Integrating renewable energy technology into the country's power supply is a great way to lessen reliance on fossil fuels and broaden the economy's base of support (Saqib, Sharif, et al., 2022;Usman & Balsalobre-Lorente, 2022).
Thirdly, several studies have examined at the dynamic links between investment levels and CO 2 emissions (Shahbaz et al., 2020;Zhao et al., 2016), demonstrating that rising investment is also a major source of CO 2 emissions.So, as the economy grows, investment in clean industries and service sectors will strengthen environmental laws and, as a result, reduce CO 2 .A recent study by Liu et al. (2017) and Ozcan (2013) reveals that investment has the potential to cut CO 2 emissions, and they propose the use of advanced clean technology acquired through F.D.I.

Data
The sample consists of panel data for 32 O.E.C.D. countries for the timeframe 1991-2020, retrieved from the International Energy Agency and O.E.C.D. databases of environmental data and indicators.In addition, we employ G.D.P. per capita, the supply of renewable energy, the development of patent technologies and gross fixed capital formation as exogenous variables, while CO 2 emissions per capita serve as the endogenous variables in our analysis.
Variable definitions are provided in Table 1.To be more precise, 'CO 2 ' denotes the annual average global CO 2 emissions per person.To determine this, we divide the total CO 2 emissions attributable to fossil fuel use by the total population.Per capita G.D.P. measures a country's economic well-being at current value and purchasing power parity (P.P.P.).R.E.N. is the renewable energy source's primary energy output in kilowatt-hours.In many earlier studies, researchers have used the I.N.O.V. definition of a country's inventive capacity as a collection of patents on environmental technologies.Investment in fixed assets (G.F.C.F.) is the difference between the value of new purchases and sales of existing assets over the course of a year.In this analysis, we adopted a logarithmic scale for all variables.

Descriptive statistics
The descriptive statistics of the six variables are shown in Table 2.As both skewed and kurtosis have high values, it is clear that the variables are not normally distributed, except for CO 2 and G.F.C.F.This means that the ordinary least squares (O.L.S.) method cannot be used.

Normal distribution test
This study makes use of a normality test procedure that was created recently by Galvao et al. (2013) in order to carry out the normality test.In order to determine whether or not the data is normal, this test uses both the individual and joint skewness and kurtosis measures, as utilisation of this normality test for panel data is very common.For a standard panel-data model presented as in Equation ( 3): The parameters are contained in the p-vector b, which does not have any constants.As is traditional, the subscript i denotes the individual, while the subscript t refers to the time, where t represents a time indicator.Error components with zero-mean in this context are the individual-specific error u i and the residual error e it : The following formula as presented in Equations ( 4)-( 7) are used to determine the skewness and kurtosis of the u i and e it separately: In order to check for skewness and kurtosis, Galvao et al. ( 2013) create statistics for both the individual-specific component and the residual component.Both of these tests can be performed separately or jointly.It is possible to formulate null hypotheses for skewness and kurtosis at the individual level when the underlying distribution is normal and can be written as: Even the symmetry statistics can be reported in a standard format, as Equations ( 8) and ( 9): Consequently, the null hypotheses for these situations are as follows, based on the normality test: : s e ¼ 0 and k e ¼ 3

Panel unit root test
The panel unit root tests based on a heterogeneous model consist of examining the significance of the findings from N independent individual tests.This testing is done in order to determine whether or not the results are statistically significant.In this particular setting, I.P.S. makes use of an average statistic, but there is another testing approach that is based on aggregating the significant values that were found in the various tests.In meta-analysis, this strategy that is based on p-values has a long and illustrious pedigree (Choi, 2001;Maddala & Wu, 1999) are two notable researchers who employed an approach that was based on Fisher (1932) type tests when conducting panel unit root testing.The statistic that was proposed by (Maddala & Wu, 1999) was defined as in Equation (10) as follows, with the very essential constraint of cross-sectional independence.
When T approaches infinity and N remains constant, the chi-square statistic has a distribution with 2 N degrees of freedom.According to Banerjee (1999), the evident simplicity of this test, in addition to its robustness to statistic choice, lag time, and sample size, make it an exceptionally viable proposition.Choi (2001) suggests a comparable standardised statistic for large N samples, such as E½À2logðp i Þ ¼ 2 and Var½À2logðp i Þ ¼ 4: This statistic represents as in Equation ( 11), the cross-sectional average of the individual p-values that have been normalised.Levin et al. (2002) suggest using the following adjusted t-statistic as presented in Equation ( 12): where the mean adjustment denoted by l Ã T and the standard deviation adjustment denoted by r Ã T are simulated for different sample sizes T.

Panel quantile regression test
In this research, a panel quantile regression model with non-additive fixed effects, which was proposed by Powell (2022) is utilised in order to study the heterogeneity effects of technological innovation under various CO 2 emission distributions.If the variables have distinct impacts at the various quantiles of the independent variable, then it is reasonable to apply quantile regression for empirical research.This is because quantile regression takes into account the distribution of the dependent variable.As a direct consequence of this, an ever-expanding corpus of research has begun to combine quantile estimates with panel data.The inclusion of the fixed effects in the mean panel regression is required in order to accurately capture the within-group variation.To estimate the quantile panel data, a number of researchers utilised a similar methodology, which takes into consideration additive fixed effects.However, the additive fixed effects are different every time a different model is applied.Powell (2022) suggests that when working with panel data, the non-additive fixed effects quantile regression be used.
Compared with the traditional fixed effects quantile model that provides lnðCO 2 Þ it À a i given D it the non-additive fixed effects quantile model provides an estimation of the distribution of lnðCO 2 Þ it given D it : D it represents the explanation variables.Powell (2022) noted that the observations at the top of the lnðCO 2 Þ it À a i distribution might be at the bottom of the lnðCO 2 Þ it distribution mentioned that the estimation might be expressed in a manner that was comparable to the cross-sectional regression.The model can be expressed in the form of the following Equation (13): where lnðCO 2 Þ i, t is the amount of CO 2 that at is produced per person in country i at year t, and b j is the parameter of interest.Except for technological innovation, we set the following control variables: G.D.P. per capita, renewable energy and investment.U Ã it is the error term.The model is linear in parameters and D 0 it bðsÞ is strictly increasing in ðsÞ: For the s th quantile of lnðCO 2 Þ i, t the quantile regression relies on the following condional restiriction presented in Equation ( 14).
The likelihood of the dependent variable is shown by Equation ( 14).With Q.R.P.D. estimation, the probability might differ between and even among individuals if the difference is orthogonal to the variables.This means that Q.R.P.D. uses a configuration where restrictions are both conditional and unconditional, as shown in Equation ( 15).
The estimation with instruments Z it ¼ ðZ i1 , . . ., Z iT Þ was also proposed by Powell (2022).However, if the variables were exogenous, as D i ¼ Z i , then the identification conditions were easily satisfied.Consequently, we make our estimations using an extended version of the method of moments.The seconds of the samples can be written as in Equation ( 16): where Z i ¼ 1 T P T i¼1 Z it using Equation ( 16) the parameter set can be expressed as B sg for all t.The parameter can then be estimated as bðsÞ ¼ argmin b2b ĝ ðbÞ Âĝ ðbÞ with weighting matrix Â: The model is estimated using the Markov Chain Monte Carlo (M.C.M.C.) optimisation method.

Normal distribution test results
The normality of the variables must be confirmed to determine if the conditional mean regression model is appropriate for the data.If sample data follows a normal distribution, conventional mean regression models can be used to analyse the effects of technological innovation and other variables on CO 2 emissions in O.E.C.D. countries.If not, conditional mean estimates would be skewed and not very reliable, which would make traditional mean regression models impossible to use.Because of this, panel quantile regression, which accounts for distribution heterogeneity, is a preferable choice in this case.Because of this, we use two normal distribution tests to make sure that our data follows a normal distribution before deciding that the panel quantile regression model is right for our needs.
Table 3 shows the findings of the normality test proposed by Galvao et al. (2013).Using this table, we may conclude that, with the exception of S e , it is appropriate to reject the null hypothesis for the other four variables.Results shows that our sample's distribution of CO 2 emissions deviates significantly from the normal distribution in both directions.The normality test for CO 2 emissions indicates that the data in sample do not follow a normal distribution.Therefore, O.L.S. techniques do not work well with our data.For this reason, the panel quantile regression technique is preferable.

Panel unit root test results
Non-stationary data may cause differences in regression results; hence the panel unit root test must be performed before empirical analysis.We utilise three panel unit root tests to ensure the stationarity of five variables, i.e., CO 2 , G.D.P., R.E.N., I.N.O.V. and G.F.C.F.These tests for the unit root were provided by Levin et al. (2002), Choi (2001) and Maddala and Wu (1999).
Table 4 displays the results of the panel unit root tests.At a significance threshold of 10%, we may conclude that practically all of the variables with the exception of I.N.O.V. are non-stationary.The first-difference sequence is stationary because the null hypothesis for all five variables at the first difference could be entirely rejected at the 1% significance level.

Panel quantile regression test results
For this reason, we will perform our empirical study using the panel quantile regression technique, which accounts for the fact that our sample variables do not have a normal distribution.Unlike traditional regression methods, which only calculate the mean.As a result, it is appropriate to provide greater details about how technological innovations effect CO 2 emissions in O.E.C.D. countries.
During the process of running the regression, we select nine different quantiles so that we can present a comprehensive view of the various quantiles.In addition, in order to facilitate a comparison between the outcomes of the conventional mean regression approach and the panel quantile regression method, we present the outcomes of the O.L.S. regression as well as the panel quantile regression in Table 5 and Figure 2. As can be seen, the influence that each of the four independent variables has on the amount of CO 2 emitted is quite varied.As seen in Table 5, technological innovation reduces CO 2 emissions, which is beneficial for O.E.C.D. countries.Nevertheless, the     effects are not uniform throughout the various quantiles and instead take the form of a W-shaped curve, as shown in Figure 2. To be more specific, the negative impact is À0.0021 at the 30th quantile, and it subsequently falls slightly at the 40th and 50th quantiles after reaching its peak at the 30th quantile.After that, the adverse effects, when measured at high quantiles, practically continued to worse.
Positive effects of the G.D.P. and negative effects of G.D.P. 2 are observed in Table 5.The results confirmed the E.K.C. hypothesis by these findings, which was also proved by (Churchill et al., 2018).The majority of member countries in the Organization for Economic Cooperation and Development are themselves developed nations.They were less dependent on manufacturing and more on the service sector.The growth of industry necessitates more energy and, consequently, produces more carbon dioxide emissions than the growth of service sector (Wang & Lin, 2016).At different quantiles, the effects of G.D.P. and G.D.P. 2 are distinct from one another.In general, when contrasted with their counterparts in the high quantiles, the beneficial effects of G.D.P. were greater for the quantiles that were lower.It is possible that this is due to the high standard of living in high-income countries.Consumption contributes significantly, on average, to the rate of economic expansion in the developed countries.In the meantime, there is a favorable correlation between consumption and quality of life.The ongoing expansion of the economy will result in significant improvements to the quality of life, including the simplification of traffic patterns and the expansion of simple access to utilities like electricity and the internet.To live a life of high quality, however, would need a greater expenditure of energy.
High-quantile countries are more vulnerable to the negative effects of renewable energy than middle-quantile and low-quantile countries.This may be due to the potential of renewable energy and its diminishing marginal effects, as noted in Saqib (2022b), which study the relationship between renewable energy and CO 2 in Asian countries.Table 5 shows I.N.O.V.'s direct impact on reducing CO 2 emissions, which varies by quantile.We investigate why technological innovation can reduce CO 2 emissions and why the direct consequences vary across O.E.C.D. members.Patents development can demonstrate the negative impacts of technological innovation on the environment.Based on the findings in Table 5, it appears that technological innovation has a W-shaped detrimental influence as shown in Figure 2. In particular, impacts kept growing from the tenth quantile through the 30 th quantile, before leveling out at the 50 th quantile.Once again, the impacts begin to rise at this point, only to level out again after the 80 th quantile.Table 5 also shows that there are adverse effects that fixed investment has on CO 2 emissions.O.E.C.D. countries' greater investment in environmental infrastructure causes fixed investment's negative effects.Song et al. (2019) shown that investments in environmental infrastructure can dramatically reduce CO 2 emissions in China.These findings are quite similar to those presented above.The data from the database maintained by the O.E.C.D. show that over the past few years, there has been a consistent rise in the amount of money spent by the government on programs designed to safeguard the environment.
There is the possibility that technical innovation will have a mitigating effect on the interaction between economic growth and renewable energy .To further support this conclusion, the relationship between technological innovation and economic growth as well as technological innovation and renewable energy demonstrates direct impacts of both variables on CO 2 emissions when the two interact with one another.The interaction effect measures the stability of the relationship between model parameters and the dependent variable.In order to evaluate the moderating effects, we conduct a second panel quantile regression, this time using the modified regression model-2 that is followed.Table 6 contains an estimation of the results obtained from model-2.Based on the results of the quantile regression shown in Table 6, we can say two things about the moderating effects of I.N.O.V.: Firstly while the effect coefficients of the I.N.O.V. Ã G.D.P. is on the negative side, CO 2 emissions and patents in particular have the potential to offset this at higher quantiles.This moderating effect might be explained by the fact that technological innovation makes a contribution to the lowering of energy intensity and, as a result, counteracts some of the beneficial effects that are caused by economic expansion (Hille & Lambernd, 2020).More and more patents, products, and processes are being developed to save energy; when put into practice, these innovations can significantly lessen energy use.If energy use per unit of G.D.P. dropped, the economy would produce less carbon dioxide.According to O.E.C.D. statistics, O.E.C.D. countries as a whole had a dramatic decrease in energy intensity between 1991 and 2020.Our rationale for the varying effects is as follows: firstly, high-quantile countries are less developed.As increased attention is paid to carbon dioxide (CO 2 ) emissions, these countries are turning to technological innovation as they industrialise.This boosts economic growth as technology progresses.Henceforth, because coefficients vary across quantiles, patents can mitigate renewable energy's effects on CO 2 emissions at low quantiles while balancing them at high quantiles.Renewable energy is minimal in high-quantile countries.Renewable energy is extremely advanced.With enough inventions, new electricity generation processes or lower production costs may be possible.Technological innovation can boost the renewable energy sector, reducing CO 2 emissions.Figure 3 shows model results.

Robustness check results
As a robustness check, we utilised different methodology in order to evaluate the consistency of the estimated parameters that were derived from the quantile regression panel  data with non-additive fixed effects.To begin, we utilised an alternative panel quantile model that is characterised by additive fixed effects for the purpose of comparison.Koenker (2004) developed this panel quantile model with additive fixed effects, which is currently widely utilised in energy and environmental investigations (Cheng et al., 2018).
Table 7 displays the conclusions drawn from the regression analysis conducted using a panel data model with additive fixed effects.The estimation findings of the robust check are compatible with those of the quantile regression panel data model, as evidenced by the regression results.As a result, we have arrived at the conclusion that the parameters of our model that were gathered via quantile regression panel data are consistent.

Conclusion
The focus of this research is on the effect that innovation in technology has on carbon dioxide (CO 2 ) emissions in countries that are members of the O.E.C.D. It is decided to use economic growth, renewable energy, and fixed investment as independent variables.Because the data are not normally distributed, a recently developed

Policy recommendations
The following is a summary of policy recommendations that have been suggested on the basis of the results and discussion, graphical trinity policy recommendation presented in Figure 4.
1.One of the major tenets of the Paris Agreement is the need for adaptation.If government and policy makers take strong action to reduce emissions, they can lessen the severity of climate change's impacts and thus the extent to which we need to adapt.Even so, climate change is already having a big impact.This makes it even more important to plan for and invest in adaptation and resilience.It is important for the government to support research and development of technology that can reduce overall energy use.The level of energy intensity is one of the ways in which technical innovation can help to moderate the effects of climate change.As a consequence of this, the government ought to foster and stimulate the implementation of this kind of technological advancement.In this manner, it would be possible to realise both the direct affects and the moderating impacts that technological innovation has. 2. To attract the necessary investments for a dramatic transition, governments need to support pro-growth structural reform policies with consistent climate policies and an investment policy environment that is well-aligned.This is necessary in order to attract the necessary investments.The most effective policy combinations to mobilise investment in low-emission infrastructure differ from country to country, including the respective contributions of public and private investment (Yang et al., 2022).In metropolitan settings, inconsistent land-use and transit planning can lock in carbon-intensive infrastructure and behavior.One of the most effective policy combinations is to have a combination of public and private investment.On the other hand in order to facilitate the adoption of new technologies, the government should do away with any obstacles in its way.Alternative energy sources are another moderating effect of technological innovation.The challenges in transferring this technology mean that renewable energy may have unintended consequences, but technological advancement may mitigate these effects.The government should remove these constraints if it is to maximise the benefits of renewable energy and technical innovation.One policy option is to make new rules about the environment, which is seen by many as a good way to speed up the implementation of new patents.3. Carbon pricing has the potential to be an effective and efficient means of encouraging low-carbon and growth-oriented actions and investments from businesses and people.To date, however, carbon prices have been very modest, especially when 'effective carbon rates' are used as a metric.These rates take into account both explicit carbon pricing and the carbon price equivalent of energy taxes.
Where carbon pricing do exist, they have had a muted and indirect effect on infrastructure investment, in part because transitional support packages and exemptions for businesses and families have diluted the price signals.The effectiveness and political acceptability of carbon pricing might also be diminished by misallocating the public revenues generated through carbon price.A better budgetary space and a more inclusive and progressive climate policy are possible with the proceeds from carbon pricing, for instance through the reduction of other taxes and the easing of the burden on the poorest households.

Figure 1 .
Figure 1.The O.E.C.D. countries' position on renewable energy and patents for environmental technology (Year: 2020).Source: The authors.

Figure 2 .
Figure 2. Technology-induced change in panel quantile regression coefficients.Source: The authors.

Figure 3 .
Figure 3. Graphical presentation of empirical findings.Source: The authors.

Figure 4 .
Figure 4. Trinity policy for climate and growth.Source: The authors.

Table 1 .
Economic growth, Renewable energy supply, Technological innovations, Investment, Technological innovations Ã Economic growth, Technological innovations Ã Renewable energy supply) Data variables and sources.

Table 4 .
Panel unit-root tests results.

Table 5 .
Panel quantile regression and O.L.S. results.

Table 6 .
Panel quantile regression and O.L.S. results (with moderating effects).

Table 7 .
Robustness results.quantile regression method is used to eliminate the possibility of biased results and provide deeper insight into the effects.At various quantiles, the empirical findings are inconsistent.In addition, a comprehensive examination of the factors that drive technological advancement is provided.In this article, we will be talking about the ways in which technological advancement has both a direct and indirect effect on carbon dioxide (CO 2 ) emissions in O.E.C.D. countries.The main conclusions of this study are listed as follows: (1) The development of new technologies leads to a reduction in CO 2 emissions; (2) Through the reduction in energy intensity, technological innovation has the potential to mitigate the beneficial effects of economic growth on CO 2 emissions; (3) The adoption of new technologies has the potential to amplify the detrimental effects of renewable energy sources on CO 2 emissions by boosting the growth of the renewable energy sector in countries with low quantiles; and (4) In O.E.C.D. countries, the E.K.C. hypothesis is found to be valid.
Note: Ã and ÃÃ indicate the significance level at 1% and 5%, respectively.Standard error values are indicated via parenthesis.Source: Authors' Estimation.panel