Impact of economic policy uncertainty on CO₂ emissions: evidence from top ten carbon emitter countries

Abstract

Over the last few decades, economic policy uncertainty (EPU) has surged across the globe. Furthermore, EPU affects economic activities, which may also generate strong CO₂ emissions. The goal of this study is to explore the impact of EPU (measured by the world uncertainty index) on CO₂ emissions in the case of the top ten carbon emitter countries, spanning the period 1990 to 2015. The findings from the PMG-ARDL modelling approach document that the world uncertainty index (WUI) affects CO₂ emissions in both the short and the long run. In the short run, a 1% increase in WUI mitigates CO₂ emissions by 0.11%, while a 1% rise in WUI escalates CO₂ emissions by 0.12% in the long run. The findings could have some substantial practical effects on economic policies through which policy makers try to shrink any uncertainty by organizing and participating in international summits and treaties. In addition, international organizations could also launch certain programs to shrink uncertainties associated with economic policy. Finally, these countries should introduce innovation, renewable energy, and enforce alternative technologies that are environment friendly. Overall, governments must provide strong tax exemptions on the use of clean energy, while R&D budgets should also expand.

Introduction

Over the last few decades, concerns about economic policy uncertainty (EPU) have escalated across the globe. In addition, the country reports of IMF (International Monetary Fund) conclude that EPU is one of the main reasons behind meagre economic growth over the last few years. Moreover, there is plethora of studies that probe the effect of EPU on different economic indicators, such as economic growth (Baker et al. 2016; Sahinoz and Erdogan Cosar 2018), investment (Kang et al. 2014), stock markets (Rehman and Apergis 2019), and energy prices (Kang and Ratti 2013).

On the top of the economic effects of EPU, it may also have environmental effects. EPU may prompt producers to employ traditional and environment unfriendly means of production, which increase CO₂ emissions. Moreover, EPU could affect consumption and investments, which in turn plunge CO₂ emissions. Furthermore, decreases in R&D, innovations, and renewable energy consumption due to high EPU could increase CO₂ emissions. Hence, the relationship between EPU and CO₂ emissions should be explored in order to propose the policies related to environmental degradation.

There are several studies in the literature that explores the effect of EPU on CO₂ emissions. Jiang et al. (2019) conclude that high EPU affects the decision-making of economic agents, increasing CO₂ emissions in the USA. In addition, Adedoyin and Zakari (2020) conclude that EPU decreases energy consumption and economic growth, plunging CO₂ emissions in the short run. Danish et al. (2020) note that EPU increases energy consumption, which surges CO₂ emissions in both short and long run. Recently, Wang et al. (2020) and Adams et al. (2020) also note that EPU escalates CO₂ emissions. Therefore, EPU can either increase or decrease CO₂ emissions (environmental degradation).

Based on the above background, the goal of this study is to investigate the effect of EPU on CO₂ emissions in top ten carbon emitter countries, namely, China, the USA, India, Russia, Japan, Germany, Iran, Saudi Arabia, South Korea, and Canada. The study contributes to the literature in three ways. First, there is limited literature that investigates the impact of EPU on CO₂ emissions. The current study fills this gap by examining its impact on CO₂ emissions in the top ten carbon emitter countries.

Second, previous studies employ the EPU index, developed by Baker et al. (2016), as an indicator for EPU. However, there are few limitations with respect to the EPU index. EPU index just covers the uncertainty related to economic policies (monetary policy, trade policy, and fiscal policy) and does not incorporate the uncertainty related to political events.¹ Moreover, the EPU index for different countries is not calculated from the single base, which creates the issues of accuracy, reliability, and ideological bias.² To overcome these limitations, Ahir et al. (2019) develop the world uncertainty index (WUI) for 143 countries. It is calculated on the basis of Economist Intelligence Unit (EIU) country reports. Furthermore, WUI is superior to EPU index as it is calculated from the single base (i.e., EIU reports) and incorporates both economic and political developments (events) in a country. This study, therefore, employs WUI as a proxy for EPU and examines the impact of WUI on CO₂ emissions.

Third, the prior literature on determinants of CO₂ emissions extensively employed first generation panel data methods, which do not incorporate the issues of cross-sectional dependence and heterogeneity. Also, these aforementioned issues may lead to unreliable results; therefore, present study employs second-generation panel data methods to overcome the issues of cross-sectional dependence and heterogeneity.

Literature on the determinants of CO₂ emissions

This section reports the prior literature on the determinants of CO₂ emissions. The previous studies highlight several economic and non-economic influencing factors of CO₂ emissions. However, economic growth is considered as one of the major determinants of CO₂ emissions (Apergis and Payne 2010). In growth-emissions nexus, environmental Kuznets curve (EKC) has widely been explored, which is inverted U-shaped relationship between income and environmental degradation (Apergis and Ozturk 2015; Aslan et al. 2018; Narayan and Narayan 2010; Murshed et al. 2020). In addition to this, energy consumption is also regarded as one of the key determinants of CO₂ emissions (Adedoyin and Bekun 2020; Zhang and Lin 2012). Also, several studies disaggregate energy consumption (i.e., renewable and non-renewable energy consumption) and highlight that non-renewable energy escalates CO₂ emissions, whereas renewable energy consumption mitigates CO₂ emissions (Alola et al. 2019; Baloch et al. 2019; Dogan and Seker 2016; Dogan and Ozturk 2017; Zaidi et al. 2018). Similarly, previous studies also note that natural resources are also driving factors of CO₂ emissions (Bekun et al. 2019; Danish et al. 2019; Joshua and Bekun 2020). Additionally, prior literature also reveals that trade surges the level of CO₂ emissions (Farhani and Ozturk 2015; Shahbaz et al. 2013). Further, there are several studies that note globalization and urbanization as one of the important determinants of CO₂ emissions (Destek 2020; Sadorsky 2014; Shahbaz et al. 2017). In addition to this, population of the country also contributes to CO₂ emissions (Begum et al. 2015; Mohsin et al. 2019). Moreover, economic policies (e.g., monetary policy and fiscal policy) also affect the level of CO₂ emissions (Ullah et al. 2020a).

There are several studies that discern the determinants of CO₂ emissions in top emitter countries. For instance, Amin et al. (2020) employ quantile regression approach to explore the impact of financial development on CO₂ emissions in top ten emitter countries. The study highlights that EKC exists for top ten emitters, and financial development also escalates CO₂ emissions. Ertugrul et al. (2016) explore that income, energy consumption, and trade are main determinants of CO₂ emissions in top ten emitters from developing countries. Mohmmed et al. (2019) report that income, population, human development index (HDI), and energy intensity are the driving factors of CO₂ emissions in top ten emitter countries. Similarly, Nejat et al. (2015) report that economic growth, population, and urbanization are the main causes of high level of CO₂ emissions in top ten carbon emitter countries. Recently, Ullah et al. (2020b) highlight that there is asymmetric effect of oil prices on CO₂ emissions in top ten emitter countries. Fatima et al. (2020) highlight that income, non-renewable energy, and renewable energy consumption are the major driving factors in top eight emitter countries. Li and Jiang (2020) explore research and development as one of the prime determinants in top six carbon emitter countries. In addition to this, Ali et al. (2020) highlight that eco-innovation, trade, and renewable energy effect CO₂ emissions in top ten emitter countries.

Parallel to this, there are several studies that explore the relationship between economic policy uncertainty (EPU) and CO₂ emissions. For instance, Jiang et al. (2019) employ granger causality in quantiles and report that EPU escalates CO₂ emissions in the USA. Similarly, Adedoyin and Zakari (2020) conclude that EPU decreases CO₂ emissions in the short run, whereas it escalates them in the long run. Danish et al. (2020) note that EPU increases energy consumption, which surges CO₂ emissions in the USA. Pirgaip and Dinçergök (2020) also report that EPU increases CO₂ emissions in the G7 countries. Recently, Adams et al. (2020) employ world uncertainty index (WUI), as a proxy for EPU, and explore the relationship between EPU and CO₂ emissions in countries with high geopolitical risk. The study reveals that EPU (measured by WUI) escalates CO₂ emissions. Similarly, Wang et al. (2020) also employ WUI (as a proxy for EPU) and report that EPU increases the CO₂ emissions in the USA.

Given the above discussion, this can be seen that relationship between EPU and CO₂ emissions has not been yet explored in top ten emitter countries. Moreover, there is dearth of literature that employs WUI (as a proxy for EPU) and investigates uncertainty-emissions relationship. Thus, the present study fills these gaps by probing the impact of WUI (i.e., proxy for EPU) on CO₂ emissions in top ten carbon emitter countries.

Theoretical background

This section elaborates the theoretical linkages between EPU (economic policy uncertainty) and CO₂ emissions. Jiang et al. (2019) describe that EPU effects CO₂ emissions through direct policy adjustment effect and indirect economic demand effect. Direct policy adjustment effect explains that high EPU diverts the attention of policy makers from environmental protection measures to economic stabilization measures, which increases CO₂ emissions. On the other hand, indirect economic demand effect describes that EPU alters the economic conditions and decision-making, which in turn effect energy consumption. Thus, the change in energy consumption ultimately effects CO₂ emissions.

Additionally, prior literature related to EPU highlights that EPU effects FDI, investment, trade, stock markets, economic development, innovations, and oil prices (Arouri et al. 2016; Canh et al. 2020; Kang et al. 2014; Sun et al. 2020; Tam 2018; Xu 2020). On the other hand, several studies report that FDI, investment, trade, stock market, economic development, innovations, and oil prices affect CO₂ emissions (Alam et al. 2020; Danish et al. 2019; Hashmi and Alam 2019; Omri et al. 2014; Sadorsky 2009; Salahuddin et al. 2018; Shahbaz et al. 2013). Therefore, this can be concluded that EPU effects CO₂ emissions through FDI, investment, trade, oil prices, etc.

Recently, Wang et al. (2020) conclude that EPU effects CO₂ emissions through two channels (i.e., consumption effect and investment effect). Consumption effect explains that EPU plunges both energy consumption and pollution-intensive products’ consumption, which in turn mitigates CO₂ emissions. On the contrary, investment effect concludes that EPU discourages the investment in R&D (research and development), renewable energy, and innovations. Meanwhile, the reduction in investment escalates CO₂ emissions. Therefore, EPU can either increase or decrease CO₂ emissions.

Methodology

Model

The analysis is principally based on the underlying intuition of the STIRPAT approach presented by Dietz and Rosa (1994). In fact, the STIRPAT model has been taken from the IPAT model, developed by Ehrlich and Holdren (1971), which probes the effects of socioeconomic determinants of environmental quality. In lieu of the fact that the IPAT has various advantages, there are also a few drawbacks of this approach. York et al. (2003) note that the hypothesis testing cannot be applied on the IPAT model because of its mathematical form. Next, the model assumes fixed proportionality across the independent variables, which is not realistically valid. In addition, the IPAT approach cannot make a distinction between the relative eminences of each factor. To overcome these drawbacks, the STIRPAT model remedies them and investigates the stochastic impact of population, affluence, and technology on environmental quality. The standard form of STIRPAT model is expressed as follows:

$$ \log \left({\mathrm{CO}}_{2,\mathrm{it}}\right)=\upvarphi {\mathrm{P}}_{\mathrm{it}}^{\upalpha}{\mathrm{A}}_{\mathrm{it}}^{\upbeta}{\mathrm{T}}_{\mathrm{it}}^{\upgamma}{\upvarepsilon}_{\mathrm{it}} $$

(1)

Moreover, we transform all variables into their logarithmic form to control heterogeneity (Farhani et al. 2014). The new model yields:

$$ \log \left({\mathrm{CO}}_{2,\mathrm{it}}\right)=\upvarphi +\upalpha \left(\log {\mathrm{P}}_{\mathrm{it}}\right)+\upbeta \left(\log {\mathrm{A}}_{\mathrm{it}}\right)+\upgamma \left(\log {\mathrm{T}}_{\mathrm{it}}\right)+{\upvarepsilon}_{\mathrm{it}} $$

(2)

In (2), φ is the intercept, whereas ε_it is the error term. Additionally, α, β, and γ are coefficients, with i and t representing cross-section and time, respectively. The empirical model used is reported in Eq. (3):

$$ \mathrm{l}{\mathrm{ogCO}}_{2,\mathrm{it}}={\upbeta}_0+{\upbeta}_1{\mathrm{logGDP}}_{\mathrm{it}}+{\upbeta}_2\mathrm{logGDP}{2}_{\mathrm{it}}+{\upbeta}_3\log {\mathrm{ENE}}_{\mathrm{it}}+{\upbeta}_4\log {\mathrm{POP}}_{\mathrm{it}}+{\upbeta}_5\mathrm{logW}{\mathrm{UI}}_{\mathrm{it}}+\upalpha \mathrm{i}+{\upvarepsilon}_{\mathrm{it}} $$

(3)

CO₂ denotes carbon dioxide emissions, GDP is GDP per capita, and GDP2 is square of GDP. Additionally, ENE denotes energy consumption, whereas POP is total population. WUI is the world uncertainty index (which is used as a proxy for economic policy uncertainty), ε_it shows the error term, and α_i denotes country fixed effects. Further, β₀ is intercept, and β₁, β₂, β₃, β₄, and β₅ are slope coefficients.

In previous studies on the determinants of CO₂ emissions for top emitters, GDP, energy consumption, and population have been extensively employed as major driving factor of CO₂ (Fatima et al. 2020; Mohmmed et al. 2019). Therefore, we also use these aforementioned variables as control variables in the present study. We incorporate GDP and GDP2 to examine the existence of EKC (environmental Kuznets curve); therefore, the expected sign of GDP and GDP2 is positive and negative respectively (Apergis and Ozturk 2015). Next, energy consumption (e.g., fossil fuel energy) is considered as a prime reason of CO₂ emissions. Therefore, the envisaged sign of ENE is positive, i.e., an increase in ENE yields higher CO₂ emissions (Danish et al. 2020). Further, high population growth exerts pressure on demand for goods and services, which escalates CO₂ emissions. Therefore, population and CO₂ emissions are expected to be positively correlated (Alola et al. 2020). Moreover, the envisaged sign of WUI is positive, implying that WUI escalates CO₂ emissions (Adams et al. 2020; Wang et al. 2020).

Next, to the best of our knowledge, there is no study that employs STIRPAT model to explore the impact of economic policy uncertainty (EPU) on CO₂ emissions. Prior studies, for instance, Adams et al. (2020), Danish et al. (2020), and Wang et al. (2020) use well-known EKC model to probe the uncertainty-emissions relationship. This motivates the current study to employ STIRPAT model and investigate the uncertainty-emissions relationship.

Methodology

As the objective is to discern the dynamic relationship between WUI and CO₂ emissions, the study employs the panel ARDL model developed by Pesaran and Smith (1995) and Pesaran et al. (1999). Pesaran et al. (1999) argue that panel ARDL approach is relatively efficient in long panel time series data. The methodology generates both short- and long-term coefficients, while it allows different lags for the dependent and independent variables. Further, the methodology is applicable if the variables are integrated at different orders (I(1) and/or I(0)). The panel ARDL model is reported in Eq. (4).

$$ \log {\mathrm{CO}}_{2,\mathrm{it}}={\sum}_{j=1}^p{\uptau}_{\mathrm{it}}\log {\mathrm{CO}}_{2,i,t-j}+{\sum}_{j=0}^q{X}_{i,t-j}{\theta}_{\mathrm{ij}}+{\rho}_i+{\varepsilon}_{\mathrm{it}} $$

(4)

CO₂ indicates carbon dioxide emissions, whereas X is the vector of all independent controls (population, energy, and GDP). Moreover, τ and θ are the coefficients to be estimated, ρ_i indicates the cross-sectional effects, whereas ε_it is the error term. Subscripts i and t, respectively, show the cross-section and time. In addition, an error correction (ECM) model can be posted as follows:

$$ \log \Delta {\mathrm{CO}}_{2,\mathrm{it}}={\eta}_i{\mathrm{ECT}}_{\mathrm{it}}+{\sum}_{j=1}^{p-1}{\tau}_{\mathrm{ij}}\varDelta \log {\mathrm{CO}}_{2,i,t-j}+{\sum}_{j=0}^{q-1}{\varDelta X}_{i,t-j}{\alpha}_{\mathrm{ij}}+{\varepsilon}_{\mathrm{it}} $$

(5)

$$ {\mathrm{ECT}}_{i,t}=\log {\mathrm{CO}}_{2,i,t-1}-{X}_{\mathrm{it}}\theta $$

(6)

In Eqs. (5) and (6), Δ denotes the first difference, and ECT is the error correction term. Next, η_i is the short-run coefficient, whereas θ is the long-run coefficient.

However, panel ARDL has three specifications, namely, PMG (pooled mean group), MG (mean group), and DFE (dynamic fixed effect) estimator. MG estimator, developed by Pesaran and Smith (1995), renders heterogonous estimated coefficients across all cross-sections in both short run and long run. Next, PMG estimator, presented by Pesaran et al. (1999), provides homogenous parameters for all cross-sections in long run. But, PMG gives heterogonous coefficients in short run. On the contrary, DFE estimator renders homogenous parameters across all cross-sections in both short run and long run. To compare the consistency and efficiency of these three aforementioned estimators, we apply Hausman (1978) specification test.

Data

The analysis uses data for the top ten carbon emitter countries (China, the USA, India, Russia, Japan, Germany, Iran, Saudi Arabia, South Korea, and Canada), spanning the period 1990-2015. The dependent variable is CO₂ emissions (metric ton per capita), whereas the control variables are GDP per capita (constant 2010$), energy consumption (oil equivalent per capita), and total population. Further, the key independent variable is world uncertainty index (WUI), which is used as a proxy for economic policy uncertainty (EPU). WUI is available on quarterly bases; therefore, we take average of four quarters to convert the data into annual frequency. The WUI is measured by calculating the frequency of word “uncertainty” (or its variants) in EIU (economic intelligence unit) reports. Next, high value of WUI implies high EPU. Also, WUI renders three quarter moving average. For instance, 2013Q4 = (2013Q4 × 0.6) + (2013Q3 × 0.3) + (2013Q2 × 0.1)/3. However, further details are given at worlduncertaintyindex.com. Figs. 1, 2, 3, 4, 5, 6, 7, 8, 9, and 10 illustrate the WUI for top ten emitters.

Fig. 1

The WUI for China

Fig. 2

The WUI for the USA

Fig. 3

The WUI for India

Fig. 4

The WUI for Russia

Fig. 5

The WUI for Japan

Fig. 6

The WUI for Germany

Fig. 7

The WUI for Iran

Fig. 8

The WUI for Saudi Arabia

Fig. 9

The WUI for Korea

Fig. 10

The WUI for Canada

In Figs. 1, 2, 3, 4, 5, 6, 7, 8, 9, and 10, blue line is the actual WUI whereas black line is the trend line. As can be seen in Figs. 1, 2, 3, 4, 5, 6, 7, 8, 9, and 10, WUI increases over the time in most of the top ten emitter countries. However, on average, WUI plunges in Canada and India. The variables used are reported in Table 1, whereas Table 2 renders some descriptive statistics.

Table 1 Summary of data

Table 2 Descriptive statistics

All data series are negatively skewed except POP, which is positively skewed. Jarque-Bera test reports that all series are not normally distributed.

Results and discussion

Unit root tests

To restrain from any spurious regression results, this first part of the empirical analysis employs the CIPS (cross-sectionally augmented IPS) unit root test by Pesaran (2007) to examine stationarity in the data. The CIPSunit root test incorporates the issues of cross-sectional dependence and heterogeneity; therefore, it is superior to other panel unit root tests (e.g., Levin et al. (2002) test and augmented dickey fuller-Fisher chi-square test). The findings from the CIPS test are reported in Table 3.

Table 3 Results from unit root test

The findings clearly highlight that all series are non-stationary in their levels, as we could not reject the null hypothesis of a unit root at the 1% significance level. In contrast, the null of a unit root is rejected in their first differences.

Westerlund (2007) co-integration test

We also employ Westerlund (2007) co-integration test to examine the long-run relationship between dependent and independent variables of our study. Westerlund (2007) test renders reliable results even in the presence of cross-sectional dependence and heterogeneity (Dogan et al. 2020). This advantage of the test compels to employ Westerlund (2007) test. The findings from the test are reported in Table 4.

Table 4 Results from Westerlund (2007) test

As can be seen, the null hypothesis of no co-integration can be rejected. Therefore, there exists a long-run relationship across carbon emissions and selected independent variables (i.e., GDP, WUI, POP, and ENE,).

Panel ARDL results

The present study employs Hausman (1978) test to discern the appropriate specification of panel ARDL model. The findings from the test are reported in Table 5.

Table 5 Results from Hausman test

As can be seen in Table 5, we fail to reject all null hypotheses. Therefore, in our case, PMG-ARDL specification is appropriate. The findings from the PMG-ARDL model are reported in Table 6; they illustrate the impact of WUI on CO₂ emissions in both the short and long run. The short-run estimates are presented with one lag, since higher lags turned out to be statistically insignificant.

Table 6 Results from the PMG-ARDL model

More specifically, they highlight that in the short run, the coefficient of WUI is negative and statistically significant. A 1% increase in WUI plunges CO₂ emissions by 0.11%, or a 1% increase in WUI decreases carbon emissions by 0.93 metric tons per capita. In addition, coefficient of GDP and GDP2 is positive and negative, respectively. Moreover, the aforementioned coefficients are also statistically significant; thus, we validate the existence of environmental Kuznets curve. Also, a 1% increase in ENE escalates CO₂ emissions by 0.31%. In addition, we do not report all those coefficients which are statistically insignificant (e.g., POP). The ECT is also negative and statistically significant, implying that any deviation from the long-run equilibrium is corrected by 76% each year.

In the long run, the coefficient for WUI is positive and statistically significant. The value of WUI is 0.12, indicating that a 1% increase in WUI increases CO₂ emissions by 0.12% or that 1% increase in WUI compels CO₂ emissions to increase by 1.01 metric tons per capita. In addition, the coefficients for POP and ENE are positive and statistically significant, indicating that increases in population and energy consumption also escalate CO₂ emissions. Furthermore, coefficient of GDP and GDP2 is positive and negative, respectively. Thus, we conclude that EKC does exist in top 10 carbon emitter countries.

Discussion

The findings reveal that WUI affects CO₂ emissions in both the short and long run. In the short run, WUI ameliorates environmental quality, as it plunges CO₂ emissions. There are two potential channels behind this result. First, high WUI (EU) may discourage energy consumption, investments at the firm level, firm’s earnings and cash flows, and tourism and GDP growth (Ali 2001; Kang et al. 2014; Adams et al. 2018; Akadiri et al. 2020), which mitigate CO₂ emissions (Danish et al. 2019; Dogan and Ozturk 2017). Second, high WUI may affect the decision-making of economic agents, which further plunges CO₂ emissions. Moreover, we also report that consumption effect is dominant in short run. These findings are in line with the conclusion of Adedoyin and Zakari (2020). The US-China trade war has increased economic policy uncertainty, which affect the decision-making about economic activities (FDI and trade). The ambiguity and inconsistency in decision-making also affect CO₂ emissions.

By contrast, in the long run, WUI increases CO₂ emissions, implying that WUI contributes to environmental degradation. There are two possible mechanisms behind this finding. First, WUI may discourage R&D, innovations, and renewable energy consumption, which escalate CO₂ emissions. The political tensions of the USA with other countries (e.g., China, Iran, and Korea) compel the USA to cut expenditures on R&D, innovations, and investments in renewable energy. Recently, President Trump cut 21% in R&D expenditure, aggravating CO₂ emissions. Second, WUI also prompts producers to employ traditional (outdated) and environment unfriendly means of production (machines that use oil as an input, while they have a low capital to output ratio), which surge CO₂ emissions (Jiang et al. 2019). Further, we conclude that investment effect is dominant in long run. These findings are backed by the conclusion of Pirgaip and Dinçergök (2020), Adams et al. (2020), and Wang et al. (2020). However, economic growth, energy consumption, and population are also responsible for environmental degradation, as they increase CO₂ emissions.

Conclusion

In the last few decades, the economic policy uncertainty (EPU) has experienced profound upsurge. In addition to the economic effects of EPU, there are also environmental effects as well. On this basis, the present study explored the impact of EPU (measured by world uncertainty index) on CO₂ emissions for the top ten carbon emitter countries. The findings from the PMG-ARDL modelling approach documented that WUI (world uncertainty index) affected CO₂ emissions in both the short and long run.

Based on these findings, a few policy implications can be deduced. First, economic policies should be very clear and transparent, with government officials trying to shrink any policy uncertainty through international summits and treaties. Second, the international organizations like UNO, WTO, and the World Bank should launch programs to shrink the economic policy uncertainties. Third, in the short run, curbing CO₂ emissions in the top ten carbon emitter countries is also possible at the cost of WUI. Therefore, if these countries crave to mitigate environmental pollution and WUI simultaneously, they should introduce innovation, renewable energy, and enforcement alternative technologies that would be employment friendly. Governments are urged to give tax exemptions on the use of clean energy, while R&D budgets should increase. In addition, grants and projects on innovations and clean energy technologies should be awarded, while subsidies should be provided on the import of renewable energy products.