Women in power: the role of gender in renewable energy policymaking

Do female policymakers encourage the production of renewable energy compared to their male counterparts? Using instrumental variables, we conduct a cross-country analysis of 39 high-income countries for the years 1997–2020 using quota laws and women’s suffrage as instruments for women’s participation in the parliament. We find that a 1 percentage point increase in the proportion of women in the legislature increases renewable energy production by 1.54 percentage points. This study suggests that fostering policies that boost women’s participation in policy-making positions is beneficial, especially when considering the positive spillover to other countries.


Introduction
Recently, there has been an increase in the representation of females in the political field.In 2019, the proportion of women in the national legislature worldwide was 24.64% (WBI.a2020).Research has shown that different policy outcomes emerge in a society depending on the gender of the legislators; women favor their gender the most (Besley and Case 2003, Chattopadhyay and Duflo 2004, Clots-Figueras 2011, Weeks 2019).Research also shows that women choose lower risk levels than men and know and care more about the environment (McCright 2010, Xiao and McCright 2012, Funk and Gathmann 2015).Additionally, female legislators provide more public goods and support environmental policies (Duflo et al 2004, Norgaard and York 2005, Fredriksson and Wang 2011, Pearl-Martinez 2014, Mavisakalyan and Tarverdi 2019).There is evidence showing that government is among the main contributors to the rapid deployment of renewable energy around the world through the introduction of different laws and policies (Adelaja et al 2010, Johnstone et al 2010, Shrimali and Kniefel 2011, Kilinc-Ata 2016, Crossley 2019).Therefore, legislator's gender influences the implementation of environmental policies.
This study investigates whether the legislator's gender has an impact on the production of renewable energy.We conduct a cross-country analysis of 39 high-income countries for the years 1997-2020.The adoption of renewable energy and the election of women to offices have an unclear causal relationship.Do female legislators pass more laws related to environmental protection, or are countries more likely to elect female legislators intrinsically more progressive?To circumvent this issue, we create a panel data with country and year fixed effects and instrument the endogenous variable, that is, the percentage of female legislators.We use passing a quota law and the number of years since women's suffrage in the country as instruments for the analysis.The results support the hypothesis that females in the legislature encourage the production of renewable energy.Using two different instruments, we find that a 1 percentage point increase in the proportion of women in parliament increases the production of renewable energy by 1.54 percentage points, after controlling for a rich set of variables.
To allow for serial correlation within countries, we cluster the error term at the country level, reducing the sample to 39 clusters.Consequently, we adjust the standard errors using the wild bootstrap (Cameron et al 2008).Our results remain unchanged after expanding the analysis to include Brazil, China, India, Indonesia, Mexico, the Philippines, South Africa, Thailand, and Turkey.These countries account for 96.11% of all renewable energy produced worldwide without hydro-power (Ritchie 2017).We test the instruments separately and find the results are between 1.4 and 2.6 percentage points.Additionally, we randomize the quota law implementation to control for potential placebo effects.We further test the robustness of our estimates by excluding each quota-adopting country sequentially.In each case, the results remain statistically significant and closely aligned with the initial estimation.
This study contributes to the literature in several ways.First, it contributes to the literature on the impact of female legislators on the large-scale production of renewable energy, a public good (Pearl-Martinez 2014).Second, decarbonizing the economy through large-scale renewable energy production is key to mitigating climate change and can be performed more rapidly than in other sectors (Pachauri et al 2014).Thus, we show evidence that promoting the election of females in the legislature could produce and accelerate this positive externality.Nevertheless, caution must be exercised if these results are extrapolated to other countries.High-income countries are intrinsically different from middle and low-income countries.Finally, our project extends the literature on cross-country analyses that link the presence of females in the legislature with more humanitarian military interventions (Shea and Christian 2017), stricter climate change policies (Mavisakalyan and Tarverdi 2019), lower CO 2 emissions (Ergas and York 2012), more protected areas (Pearl-Martinez 2014), more environmental agreements (Norgaard and York 2005), higher health-expenditure by governments Clayton and Zetterberg (2018), more childcare expenditure policies (Weeks 2019), and less corruption and bribe acceptance (Dollar et al 2001, Swamy et al 2001). 3 The remainder of this paper is organized as follows.Sections 2 and 3 describe the relationship between the percentage of female legislators and renewable energy production.Section 4 presents the data and descriptive statistics.Section 5 describes the econometric analysis.Sections 6 and 7 present the estimation results and robustness checks, respectively.Finally, we present our conclusions in section 8, adding possible future research opportunities.

Energy and the government
Fossil fuel combustion and industrial processes accounted for 78% of the total CO 2 emissions for the period 2000-2010 (Pachauri et al 2014).The Renewable Global Status Report (Raturi 2019) established that 74 countries committed to having some percentage of their energy production based solely on renewable sources.
Figure 1 illustrates the evolution of (average) renewable energy production over time for all the countries in our sample.In high-income countries, renewable energy production has increased steadily since the late nineties 4 .
Several studies using different samples, time spans, and policies conclude that governments around the world, pushing for regulatory policies and fiscal incentives, are the main drivers for the rapid increase in renewable energy production in the world.Regulatory policies include feed-in-tariffs (FITs), electric quota obligations (also known as renewable portfolio standards), net metering, and tendering (Raturi 2019).Fiscal policies include investments, tax incentives, and public financing, such as public investments, loans, grants, and capital subsidies.For example, Crossley (2019) gathered legislation from 113 countries that promote renewable energy production.The author found that countries have passed these laws for several reasons.For example, 37 countries state as an objective in their renewable energy laws to promote sustainable development; 35 countries state as an objective the reduction of fossil fuel use and the reduction of oil dependency from other countries (Huang et al 2007); 55 countries declare as an objective the protection of the environment; and 24 countries have laws supporting the expansion of the renewable energy production industry per se.Johnstone et al (2010) conducted a panel data analysis of 25 OECD countries over the period 1978-2003.The authors analyzed the effect of several public policies on renewable energy innovation, using patent counts as a proxy for innovation 5 .They concluded that public policy is a major contributor to renewable energy innovation.Furthermore, Adelaja et al (2010) and Kilinc-Ata (2016) reached the same conclusion; Adelaja et al (2010) studied the wind energy market in the United States, while Kilinc-Ata (2016) conducted a cross-country analysis for 27 countries of the EU and the United States.Similarly, Shrimali and Kniefel (2011) showed that different policies (RPS, State Government Green Power Purchasing, and Clean Energy Fund) play an important role in the deployment of renewable energy production in the United States.
In 2019, 43 countries used tax credits as a stimulus, and 101 countries used public financing (Raturi 2019).By conducting a cross-country analysis of 27 countries in the EU and the United States for a span of

Previous literature on female leaders
As remarked by Hessami and da Fonseca (2020), the literature has moved from the idea of politicians 'just' reflecting the median voter preferences to continue in their seats, to a more accepted idea that some characteristics of the politician, such as gender and race, matter when implementing policies.In the same vein, using League of Conservation Voters data for the period 1970-1995, Fredriksson and Wang (2011) show that voters do not 'affect' environmental policies pushing them to the middle.Instead, voters 'elect' politicians who are aligned with their preferences and concerns, opposing the median voter theory.Research has found that women are more likely than men to provide public goods, favor their own gender, and push for policies that represent their own gender preferences the most (Chattopadhyay and Duflo 2004, Clots-Figueras 2011, Weeks 2019).For example, Besley and Case (2003) analyze whether women legislators prioritize policies different from those of men in the USA and find that women in power positions pass more laws regarding family assistance and child support.According to the IPCC Report 2022 (IPCC 2022), women are less likely to be climate change deniers in comparison to men.The report also finds a positive relationship between higher female political participation and stringency in climate policies.
A broad body of literature finds that female legislators seek to protect the environment.For example, in the Human Development Report, Pearl-Martinez (2014) used a sample of 90 countries and found that countries with a higher percentage of women in parliament are correlated with more protected land areas.Mavisakalyan and Tarverdi (2019) use a panel of 91 countries for the period 2005-2012 and find that more female representation in parliament leads to stricter climate change policies and a reduction in CO 2 emissions.Norgaard and York (2005) studied 130 countries and found that more women in parliament are more likely to ratify environmental treaties.Similarly, Fredriksson and Wang (2011), using The League of Conservation Voters, found that female legislators push for stricter environmental policies.
Finally, research has shown that women care more about the environment.McCright (2010) used the Gallup survey from 2001 to 2008 and found that women manifest more concern and have more knowledge about climate change than men do.Similarly, Xiao and McCright (2012), using the Gallup survey, found that women care more about health problems caused by changes in the environment.The authors showed that this greater concern is due to women having different risk perceptions than men do.Moreover, Funk and Gathmann (2015) studied the differences between men's and women's preferences by analyzing how their votes differ in several topics in Switzerland.The authors found that women are 10 percentage points more likely to spend money on policies that protect the environment than men.Finally, recent concerns of women toward the environment have also increased over time.Using the 'World Values Survey' (Inglehart et al 2018A, 2018B), we find that women worldwide have increased their concern over environmental pollution and the degree of responsibility assigned to the government.Therefore, it is possible that female legislators are trying to protect the environment to advance the interests of the female population.

Data and descriptive statistics
We use panel data for a sub-sample of high-income countries for the years 1997-2020.Table A2 in the appendix shows the list of all the countries in our sample.We focus on the high-income countries as categorized by the World Bank.These countries account for 56.19% of total renewable energy production without hydropower (Ritchie 2017).Adding Brazil, China, and India accounts for 91.26% of the total renewable energy production (Ritchie 2017).Thus, we include these countries as a robustness check.
Our variable of interest is renewable electricity production without hydropower and biomass.The data was collected from the IEA (IEA 2020).We define this variable as the proportion of electricity generated by renewable plants to total electricity production.Renewable plants include geothermal, solar, tidal, and wind plants.
We exclude hydropower following Verdolini et al (2018) and Popp et al (2011), as they argue that it is a mature renewable energy source, in which most deployment has already occurred.Nevertheless, we add hydropower to the main specification (equation ( 3)) as a robustness check.We also add the production of biomass and biofuels plants as a robustness check.
The main independent variable is the proportion of women in parliament.The World Bank defines this variable as 'the percentage of parliamentary seats in a single or lower chamber held by women' (WBI.a2020). 6The proportion of women in national parliaments in the world in 1997, considering single or lower chambers, was 11.69%, in 2019, it increased to 24.64% (WBI.a2020).During that period, the proportion of women in the single or lower chamber increased by 83.25% in high-income countries.The evolution of seats held by females in parliament for our sample is presented in figure 2.
Despite gaining more seats over time, the total participation rate of women remains fairly low.Consequently, several countries have implemented gender quota laws to reduce this gap.In our sample, 15 countries passed a gender quota law (see table A3 in the appendix for further details).We use a dummy for This table shows the year when high-income countries adopted a quota law.Source: ECLAC (2020), IDEA (2020).For more information, refer to table A3 in the appendix.when a quota law was passed in the lower or single house as an instrument (ECLAC 2020, IDEA 2020).We consider only the quotas that specify a percentage of seats for females.Table 1 shows when a quota law was adopted.
Data for the 'years since women's suffrage' is collected from historic newspapers (Historic 2020), which refers to the years when women were allowed to vote nationally.This variable is used as a second instrument.The instrument was constructed as follows.First, we obtained the year in which suffrage was expanded to women for each country in the sample (refer to table A4 in the appendix for more information).Second, we compute the difference between the years of the data (1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015) minus the year suffrage was expanded to women.
We also use the (logarithm of the) GDP per capita at constant 2015 dollars as another control (WB.c 2020).In addition, we use Freedom House's Civil Liberties Index.This index ranges from 1 to 7, where 7 represents the lowest level of freedom, and 1, is the highest.Thus, an increase in the index reflects a decrease in civil liberties (House 2020).
Furthermore, we add the political orientation (right-wing, left-wing, or centrist) of the party that wins the election (Cruz et al 2016).The election-winning party is categorized as right-wing if the party is self-defined as conservative, Christian democratic, or right-wing 7 .A left-wing party is designated as such if it considers itself socialist, social-democratic, communist, or left-wing.Finally, a party is considered centrist if it is self-defined as such.Parties that do not conform to the previous categories are ruled out.We transform the categorical variables into dummies.Lastly, we add another specification on whether the system is parliamentary or presidential (CIA 2020) (see appendix table A10 for more details).Specifically, we create a variable that equals one if the system is non-parliamentary.
Table 2 presents descriptive statistics of the data used in the regression analysis.On average, the percentage of renewable energy production without hydropower for the countries in our sample is 4.61%.However, there is a large discrepancy among countries, with one country having a maximum of 61% of renewable production.Considering renewable energy with hydropower and biomass, the average production increases to 28%.The percentage of females in the legislature is 23%.Even though female representation has increased gradually over time, as shown in figure 2, it is still far from achieving gender parity.

Methodology
Our econometric specification is the following8 : where y it is the proportion of renewable energy sources in a country's overall electricity production i at time t.
The proportion of seats held by women in the national parliament of country i at time t is represented by the variable W it .X it is a set of time-variant variables linked to both the percentage of female legislators and the use of renewable energy: the logarithm of GDP per capita of country i at time t, the government of country i at time t is right-wing, centrist, or left-wing, and the civil rights index.Time invariant country fixed effects and time dummies for 1997-2020 are represented by δ i and γ t , respectively.Finally, ϵ it is the error term.
Identifying the impact of higher female participation in parliament on the level of renewable energy production is not straightforward.Are countries that favor more renewable energy more likely to elect women to the legislature because they are inherently more progressive, or do women want more renewable energy?This is complicated to analyze without conducting an experiment.The optimal identification strategy would be to randomly select some countries to assign them a certain number of women in their parliaments and then analyze their differences in renewable production outcomes.
Additionally, there could be some omitted-variable bias, such as, cultural or historical persistence.In order to try to mitigate this problem, we apply an IV approach and add year and country fixed effects.Furthermore, we control for variables that simultaneously impact both the number of female legislators in the parliament and the production of renewable energy (X it ).To achieve this objective, we employ two instruments: quota laws and years since women's suffrage.
The variable 'countries with legislated candidate quotas in the lower or single house,' is obtained from the ECLAC (2020), IDEA (2020) websites.The instrument is equal to one if the law requires that a certain number of seats be reserved for women's (Z 1it ).Table A3 in the appendix describes the quota laws for each country.We also employ 'years since women's suffrage' (Z 2it ) (Grier and Maldonado 2015, Hicks et al 2016, Mavisakalyan and Tarverdi 2019).
Thus, the first stage of our estimation is: And the second stage: For the instrument to work, it must be first relevant.In this case, the gender quota laws and the years since women's suffrage must have a positive and significant effect on the proportion of women in the parliament.Second, the instruments must comply with the exclusion restriction.This means that the instruments must be as good as randomly assigned and can only affect the percentage of renewable electricity output by the variable we want to instrument, that is, the proportion of females in parliament.
The exclusion restriction may not hold if countries that pursue diversity strongly enough to pass a quota law also care more about the environment.Nevertheless, quota laws are passed to increase diversity, which differs from progressiveness.More progressive countries want to increase the number of females in policy-making positions and increase the production of renewable energy.A quota law increases the former but not necessarily the latter.For example, Haiti passed a quota law in 2012, granting 30% of its seats to women (ECLAC 2020).Nevertheless, their Freedom House's Civil Liberties Index score for the same year was 5, meaning that it was considered only partially free (House 2020).Second, we only consider countries in which the law requires at least a percentage of their seats to be held by females.Third, women in legislative bodies have been underrepresented for decades, therefore, passing a quota law has, in principle, the sole effect of reducing gender inequality.The channel we are exploiting by using this instrument is the following: passing a quota law increases the number of females in parliament, who push for different policies than men9 .Hicks et al (2016) and Mavisakalyan and Tarverdi (2019) show that women's suffrage is correlated with having more women in parliament (a strong first stage); therefore we use this variable as one of the instruments to estimate equations ( 2) and (3).The presence of more equitable nations does not automatically imply a higher prevalence of progressive nations.While progressive countries are characterized by greater fairness, it should be noted that fairness alone does not guarantee progressiveness.This is particularly relevant when considering the historical context, as women's suffrage occurred around 1933 in our sample.The mechanism exploited is as follows: an expansion in the right to vote for women implies that women can choose according to their beliefs, shape policy outcomes, and increase their participation.However, we expect the 'years since women's suffrage' to have a highly diluted impact on current outcomes.
We cluster the sample at the country level to allow residuals to have an unrestricted correlation within countries and time (Angrist and Pischke 2008).Our sample has 39 clusters categorized as high-income by the World Bank.However, there are disparities between them.For example, table 2 illustrates that Denmark and Uruguay have more than 40% of their energy produced by renewable sources (without considering hydropower), whereas the average of the entire sample is 4.6%.Finally, 15 out of 39 countries in the treatment have passed a quota law.Hence, we have a small number of treated clusters.Furthermore, the number of clusters is less than 42 (Angrist and Pischke 2008), and they are not homogeneous.Roodman et al (2019) recommend finding p-values via bootstrap if either of these conditions holds because they lead to over-rejection.We follow Cameron et al (2008) and use a wild-bootstrap.We present bootstrap p-values in the main estimation results10 .

Results
The OLS estimation results are presented in the first column of table 4. The coefficient for the variable 'proportion of women in parliament' , including all controls, is −0.06 (0.13), which is not statistically significant.There could be some omitted variable bias negatively correlated with the variable of interest; therefore, we instrument this variable using quota laws and years since women's suffrage.
Tables 3 and 4 present the first and second-stage estimates of the main specification (equation ( 3)).This study focuses on high-income countries defined by the World Bank, which account for approximately 56% of the total worldwide production of renewable energy (without hydropower) (Ritchie 2017).
The first stage is presented in table 3, where we use quota laws and years since women's suffrage as instruments for the proportion of women in parliament.Our preferred specification is the fourth column, which includes all the controls.In this specification, the introduction of a gender quota law increases the proportion of women in parliament by 2.3 percentage points.Allowing women to vote a year earlier increases the proportion of women in the parliament by 0.6 percentage points.Both coefficients are significant at 5%.We also find that the instruments used are robust to different specifications, as shown in columns 1-4 of table 3.
According to the results of the second-stage estimation in table 4, we find that a one percentage point increase in the proportion of women in parliament increases the percentage of renewable energy production by 1.5 percentage points as shown in column 5, which is the preferred specification.The coefficient is significant at 1%, after controlling for the type of government, the degree of civil liberty, and the log of GDP per capita in 2015 constant dollars.
The majority of the controls are not statistically significant.The coefficients for the type of party and GDP have an unexpected sign, especially, considering that an increase in the former happens when the government moves further to the 'left' .Having a right-wing government has a positive (yet not significant) impact on the production of renewable energy compared to that of a left-wing government.The opposite happens with a centrist government.Regarding the Civil Liberty Index, 1 is the highest level of freedom.Thus, decreasing a unit in the Civil Liberties Index (i.e. an improvement in the Civil Liberties Index) implies a higher effect on renewable energy, although it is not significant.Finally, the fact that the logarithm of GDP per capita is negative is also counter-intuitive.However, given that we focus on high-income countries may explain this result.For example, the level of renewable energy depends on large-scale projects that may require external funds.Furthermore, despite being counter-intuitive, our results are similar to those of Dong (2012) and Yin and Powers (2010).The results from the reduced form estimation are presented in the appendix table A5.
We also perform the Sargan-Hansen test of overidentification, and we are not able to reject the null hypothesis (the p-value is equal to 0.375).Therefore, at least one of the instruments is exogenous and is not correlate with the error term.
We try both instruments separately and find that a one percentage point increase in the proportion of women in parliament increases the percentage of renewable energy production between 1.4 and 2.6 percentage points, using years since women's suffrage and passing a quota law respectively.The first stage results are presented in the appendix tables A6 and A8.The second stage is presented in tables A7 and A9, respectively.
We also test whether having more women in the legislature has a more pronounced effect on parliamentary rather than presidential systems (see table A10 for more details).Legislators under a parliamentary system may participate more actively and decisively than those under a presidential system.Specifically, we create a variable that equals one if the government is non-parliamentary.Then, we create an interaction term ('Parliamentary * Prop.women parliament') that considers the proportion of seats represented by women and non-parliamentary government.We present the point estimates in table A11, column (1).The point estimate for such interaction is not significant, so the effect of being in a non-parliamentary system is inconclusive.However, the effect of the proportion of women in the parliament is almost unchanged.

Robustness checks
First, we conduct a placebo test by making the quota law adoption random11 .The 'random' effects should not impact the production of renewable energy.The results are presented in table A13.There is no significant effect of the proportion of women in parliament on the production of renewable energy if we use the 'fake' gender quota laws as an instrument.This validates our causal interpretation.
Second, we drop each treated country at a time and run the main specification (3) at a 95% confidence level.The results are presented in table A14 in the appendix.None of the countries drive the results alone; the point estimations are close to each other and are always significant at 1% or 5%.
Furthermore, we add three more countries, Brazil, China, and India, which, together with the original sample, account for 91.26% of the global renewable electricity production without hydropower (Ritchie 2017).The coefficient of the percentage of women in parliament decreases slightly, from 1.5 to 1.4, and is significant at the 1% level.Lastly, we add nine more countries: Brazil, China, India, Indonesia, Mexico, the Philippines, South Africa, Thailand, and Turkey to the original sample.These countries account for 96.11% of the renewable energy production without hydropower worldwide (Ritchie 2017).The coefficient on the percentage of women in parliament decreases from 1.5 to 1.2 and is still significant at 1%.The results can be seen in table A15, Columns 1 and 2, respectively.
Finally, we estimate equation (3) using the proportion of renewable energy production including hydropower as the dependent variable.The coefficient of the proportion of women in parliament is 1.4, after controlling for the type of government, the degree of civil liberty, and the log of GDP per capita in 2015 constant dollars.The effect is significant at 1%.The estimation almost does not change, and this is consistent with the idea that hydropower is considered a mature energy source where most of the deployment has already occurred (Popp et al 2011, Verdolini et al 2018).The results are presented in the appendix in table A16.We also do the above analysis using the proportion of renewable energy production including biomass and results are presented in table A17 in the appendix.
This section ratifies the relevance of the instrument to the main causal findings.The introduction of a quota law increases the number of women in parliament, which potentially, through an increase in the number of policies and laws adopted (as discussed in section 3), is one mechanism for increasing the production of renewable energy.However, these results can be extrapolated only to high-income countries.Middle and low-income countries have several dissimilarities.Although having more women in policy-making positions has a positive impact, it does not necessarily translate into an increase in renewable energy production.

Conclusion
More women in policy-making positions leads to an increase in the production of renewable energy without hydropower, after controlling for several factors.These results add to the literature on cross-country analyses that link women in parliament to several outcomes.By using an IV panel with fixed effects and considering 39 high-income countries for the years 1997-2020, we reinforce the argument that women in policy-making positions support the environment by increasing renewable energy production.
By using two different instruments, passing a quota law and years since women's suffrage to control for possible omitted variable bias, we find qualitatively similar results.Women in parliament increase the production of renewable energy production by 1.54 percentage points, after controlling for the log of GDP, type of government, and Civil Liberties Index.
It has been documented that female legislators support different policies compared to male legislators.Thus, enhancing government performance necessitates a focus on increasing the participation of women, as well as promoting diversity in a broader sense.Furthermore, this study shows that having more women in parliament has a positive and significant effect on the production of renewable energy, perhaps because of an increase in the number of environmental protection-related legislations and laws proposed by women.Consequently, having more female representatives has an important effect on the decarbonization of the economy, which is a positive externality for the rest of the world.Introducing or strengthening quota legislation is a strategy to enhance representation.
Due to the inherent differences between middle and low-income countries and high-income countries, these results cannot be extrapolated to other types of countries.Therefore, although having more women in positions of influence would undoubtedly boost diversity, it would not always result in the production of more renewable energy.
A potential avenue for further investigation would involve examining diversity across countries and delving into the reasons behind the specific developmental trajectories observed in high-income countries.This would explore the presence of alternative mechanisms and how they vary across middle-income and low-income countries.Additionally, it would investigate why middle-income nations were at the forefront of the deployment of renewable energy before things slowed down.significant at 5%, after controlling for the type of government, the degree of civil liberty, and the log of GDP per capita in 2015 constant dollars.
The IV results and the first stage using 'years since women's suffrage' as an instrument can be seen in tables A9 and A8, respectively.Allowing women to vote a year earlier increases the proportion of women in the parliament by 0.59 percentage points.In the second stage, the coefficient is 1.4 after controlling for the type of government, the degree of civil liberties, and the log of GDP per capita in 2015 constant dollars.Thus, a one percentage point increase in the percentage of women in parliament increases the production of renewable energy without hydropower by 1.4 percentage points.The coefficient is significant at 1%.This table shows the random draw that we treat as the new quotas laws.These new fake quota laws are used as instruments to test whether this random effect has any impact.The IV results of the effect of women in parliament on renewable energy production are presented.Both instruments are used: passing a quota law and years since women's suffrage.In column 1, we added three more countries: Brazil, China, and India, which, along with the original sample, account for 91.26% of all the renewable energy production done worldwide.
Column 2 also includes Brazil, China, India, Indonesia, Mexico, the Philippines, South Africa, Thailand, and Turkey.This extended sample accounts for 96.11% of all renewable energy production done worldwide.Standard errors are clustered at the country level in parentheses.Wild bootstrap p-values, in square brackets.Significance levels: * * * 0.01 * * 0.05 * 0.1.The IV results of the effect of women in parliament on renewable energy production with hydropower is presented, using both instruments, passing a quota law and years since women's suffrage.Standards errors are clustered at the country level in parentheses.Wild bootstrap p-values, in square brackets.Significance levels: * * * 0.01 * * 0.05 * 0.1.The IV results of the effect of women in parliament on renewable energy production with biomass is presented, using both instruments, passing a quota law and years since women's suffrage.Standards errors are clustered at the country level in parentheses.Wild bootstrap p-values, in square brackets.Significance levels: * * * 0.01 * * 0.05 * 0.1.
In the following table, we present the point estimation results using the percentage of renewable energy with hydropower as a dependent variable.
In the following table, we present the point estimation results using the percentage of renewable energy with biomass as a dependent variable.
18 years from 1990 to 2018,Kilinc-Ata (2016) found that FITs and tax policies help the deployment of renewable energy production.Similarly,Liu et al (2019) conducted a cross-country analysis of 29 countries from 2000 to 2015 and concluded that fiscal and financial incentives, market-based instruments, policy and support, and RD&D improve renewable energy deployment.Finally,Shrimali and Jenner (2013) conducted a cross-state analysis for the United States over the years 1998-2009, finding that tax credits, among other policies, increase the production of solar photovoltaic power by reducing its costs.

Figure 2 .
Figure 2. Percentage of seats held by females for high-income countries.Source: WBI.a (2020).

Figure A1 .
Figure A1.Percentage of renewable energy production for all European countries in our sample.Source: IEA (2020).

Table 1 .
Year of quota law adoption.

Table 3 .
First stage estimation.This table shows the effects of both instruments, passing a quota law and the number of years since women's suffrage, on the percentage of women in parliament.Standards errors are clustered at the country level in parentheses.Significance levels: * * * 0.01 * * 0.05 * 0.1.

Table 4 .
IV estimation of renewable energy.

Table A1 .
Summary of cross-country and gender analysis.

Table A2 .
Countries in the sample.

Table A4 .
Years of suffrage extension to women.

Table A5 .
Reduced form estimation.The OLS results of the effects of passing a quota law on the production of renewable energy are presented.Standards errors are clustered at the country level in parentheses.Significance levels: * * * 0.01 * * 0.05 * 0.1.

Table A6 .
First stage estimation.This table shows the impact of adopting a quota law on the percentage of women in parliament.Standards errors are clustered at the country level in parentheses.Significance levels: * * * 0.01 * * 0.05 * 0.1.

Table A7 .
IV estimation of renewable energy.the IV results of the effect of women in parliament on renewable energy production, using passing a quota law as an instrument.Standard errors are clustered at the country level in parentheses.Wild bootstrap p-values in square brackets.Significance levels: * * * 0.01 * * 0.05 * 0.1.

Table A8 .
First stage estimation.This table shows the effects of the number of years since women's suffrage was allowed on the percentage of women in parliament.Standards errors are clustered at country level in parentheses.Significance levels: * * * 0.01 * * 0.05 * 0.1.

Table A9 .
IV estimation of renewable energy.

Table A10 .
Government type.table shows the type of government each country has, as defined by the CIA (2020). This

Table A11 .
IV estimation of renewable energy production.
The IV results of the effect of women in parliament on renewable energy production are presented.The type of government is included as an extra control in column (1).'Parliamentary * prop.women parliament' equals one if the government is non-parliamentary, e.g. a presidential republic, zero if it is parliamentary.Standards errors are clustered at the country level in parentheses.Significance levels: * * * 0.01 * * 0.05 * 0.1.

Table A14 .
IV estimation of renewable energy production: dropping each country at a time.The IV results of the effect of women in parliament on renewable energy production are presented.In each line, the respective country is dropped.Significance level is set at 95%.Wild bootstrap p-values, in square brackets.Significance levels: * * * 0.01 * * 0.05 * 0.1.

Table A15 .
IV estimation of renewable energy production-extended sample.

Table A16 .
IV estimation of renewable energy with hydropower.

Table A17 .
IV estimation of renewable energy with biomass.