Female legislators and forest conservation

While women are known to be more concerned about the environment than men, do these preferences translate into action when women hold political power? This article shows that in constituencies reserved for historically marginalised communities, when female politicians win close races against male candidates, there is an immediate positive effect on forest conservation efforts. In other constituencies, the benefits accumulate over time.

Qualitative evidence suggests that women are generally more concerned about the environment as compared to men. For example, the Chipko movement¹ in India, which aimed to prevent deforestation, was largely led by women. Additionally, the 2022 World Values Survey for India shows that 61% of women support environmental protection even at the cost of economic growth, as against 55% of men. However, it remains unclear whether women's environmental preferences translate into action when women hold political power. In a recent study (Bandyopadhyay et al. 2024), we examine whether the gender of legislators influences environmental conservation in India, using changes in forest cover as a measure. Forests, as major carbon sinks, have been central to multilateral climate agreements, such as COP-26, which stress the importance of forest conservation.²

Environmental degradation and climate change disproportionately affect women and children (Currie and Rossin-Slater 2013, Sekhri and Storeygard 2013, Banerjee and Maharaj 2020). Therefore, our question is relevant in the light of existing studies that have documented that female politicians are responsive to issues that are more likely to affect women and children (Chattopadhyay and Duflo 2004, Bhalotra and Clots-Figueras 2014, Bhalotra et al. 2023). However, Baskaran et al. (2023) shows that female legislators promote economic growth in their constituencies, which naturally raises an interesting puzzle about what this implies with regard to environmental conservation. This is because economic growth and environmental protection have often been viewed as conflicting goals. Hence, it would be interesting to understand whether female politicians can indeed facilitate a sustainable growth path.

Institutional background: Politicians in state assembly legislatures

Our focus is on exploring the role of female vis-à-vis male politicians elected to India’s state legislative assemblies play when it comes to forest conservation. As is well known, elections to state legislatures usually occur once every five years and the electoral process followed is the “first past the post” system for deciding the winner, who becomes a Member of the Legislative Assembly (MLA).

We focus on MLAs rather than Members of Parliament (MPs) who are elected during the federal/general elections because, prior to 1976, forests were under the "State" list of the Indian Constitution, giving state governments exclusive authority over forest conservation. After 1976, forests were moved to the "Concurrent" list, allowing both the federal and state governments to legislate on forest management. However, anecdotal evidence shows that parliamentary discussion on environmental and climate change issues is not widespread; state government legislators appear to be more engaged regarding environmental issues. This highlights the important role MLAs play in forest management. Moreover, MLAs are allocated funds (MLA Local Area Development schemes, or MLA-LADs) for constituency development, which can be used for forest conservation activities (such as through tree plantation, social forestry, farm forestry, etc.).

Measuring forest cover changes and electoral data

We use data from the Socio-Economic High-resolution Rural-Urban Geographic Platform for India (SHRUG) (Asher et al. 2021), which provides detailed information at fine geographic units (for example, towns, villages, or assembly constituencies) with consistent boundaries over time. This allows us to study the dynamic relationship between environmental conservation and political factors at a granular level.

For our analysis, we combine satellite imagery on forest cover (distinguishing between forests and other plantations) from 2000-2011 with state assembly election data at the constituency level from 1996-2007. Asher et al. (2020) note that whereas dense forests are largely geographically concentrated in India, pockets of forests are common throughout the country. Our key outcomes of interest are the rate of change of forest cover annually as well as over an MLA’s electoral term. Since we are studying the impact of politician gender on forest cover dynamics, the earliest electoral data predate the forest cover data. Delimitation considerations also mean the latest electoral data in our analysis are from 2007.³ As MLAs usually serve five-year terms, forest cover data extend up to 4-5 years beyond the latest electoral data.

Estimating the impact of female legislators on forest cover changes

A key challenge in establishing the causal impact of politician gender on forest cover changes is that comparing constituencies with female versus male legislators may not be valid. For example, constituencies electing female politicians may already be more environmentally conscious. To address this concern, we use a regression discontinuity design (RDD), which compares constituencies where a female legislator won by a ‘narrow’ margin against a male politician with those where a male legislator won similarly.⁴ The idea is that close races are likely to be quasi-random, making it possible to establish a causal relationship between legislator gender and environmental outcomes under simple assumptions.

Findings

We find that the annual expansion in the constituency-level forest cover is not significantly impacted by female politicians who won in closely contested elections against a male politician. However, in constituencies reserved for historically marginalised communities such as the Scheduled Castes/Scheduled Tribes (SC/ST), electing a female politician in a close election against a male politician resulted in a statistically significant 6 percentage points increase in yearly forest cover growth. Figure 1 illustrates the result graphically. Our finding is comparable to that of Clots-Figueras (2011), who finds that women legislators who belong to historically disadvantaged communities invest more in public goods that especially benefit women and children (such as health and early education).

Additionally, we find that constituencies that are sparsely forested⁵ – that is, those that fall within the bottom 75% of the forest cover distribution – are the main drivers of our results. Our result with regard to annual forest cover changes at the constituency level will likely translate to an annual increase in sequestration of approximately 1,584 tonnes of carbon dioxide.

Figure 1. Yearly growth in forest cover where female politicians narrowly won or lost against male politicians

Note: Plot of regression discontinuity design estimates for (a) all constituencies and (b) SC/ST reserved constituencies where outcome is yearly growth in forest cover.

When the analysis is conducted for a variety of subsamples (for example, major states, restricting the sample by excluding the outliers in the measure of forest cover) and alternative specifications, our key finding remains unchanged. We conduct numerous robustness tests to validate the accuracy of the RDD estimates, and these are found to be robust to each of these tests.

Analysis of forest cover growth across an electoral term reveals that constituencies won by female politicians in close elections, as opposed to their male counterparts, exhibit a significant increase in forest cover growth during their tenure. Notably, we found that the long-term effects on forest cover growth (that is, the effect over the entire electoral term) are not limited to the SC/ST reserved constituencies; they are observed in all constituencies. This implies that while conservation initiatives benefit reserved constituencies right away, they have a gradual positive impact on other constituencies. Electing a female politician over her male counterpart in mixed-gender close elections, however, has a more pronounced positive and significant effect on long-term forest cover growth in SC/ST reserved constituencies (19 percentage points) than that across all constituencies (12 percentage points). The results are graphically represented in Figure 2. Similar to the annual growth rate, our findings regarding the long-term growth rate are influenced by constituencies within the lowest 75% of the forest cover distribution.

Figure 2. Growth in forest cover over an electoral term where female politicians narrowly won or lost against male politicians

Note: Plot of regression discontinuity design estimates for (a) all constituencies and (b) SC/ST reserved constituencies where outcome is growth in forest cover over a legislator’s electoral term.

Finally, we investigate the possible mechanisms underlying our findings. Specifically, across all constituencies, female legislators are around six years younger than their male counterparts. We, therefore, attempt to determine whether the discrepancy in legislator age could account for our observations regarding the influence of legislator gender on the expansion of forest cover over the electoral term. We employ an age threshold of 60 years to distinguish between older and younger legislators in state assemblies. In the cohort of MLAs under 60 years of age at the time of the election, the effect of choosing a female legislator in a close mixed-gender contest on forest cover growth throughout the electoral term is substantial (27 percentage points) and statistically significant. However, in the case of MLAs aged 60 years and older, there is no significant difference identified between female and male MLAs regarding the evolution of forest cover throughout their tenure. Our findings align with those of Dahis et al. (2023) who find that younger politicians are more likely to prioritise long-term initiatives like environmental conservation.

We show that the observed characteristics of age, education, asset ownership, and the number of offences with which one has been charged, do not appear to be systematically different between female and male politicians from SC/ST reserved constituencies. Therefore, there are probably other possible explanations for the results pertaining to the annual expansion of forest cover for the SC/ST reserved constituencies. We discuss two possible mechanisms that can jointly impact our findings. One possible source of explanation is that women and men have different behaviours and preferences when it comes to environmental protection. According to existing literature, women are more risk averse, patient, and altruistic than men (Bauer and Chytilova, 2013, Crosson and Gneezy 2009, Anderoni and Vesterlund 2001). Investing in environment and protection against climate-related shocks is similar to an intergenerational transfer that is guided by altruistic actions. These possible behavioural variations are unlikely to be uniform across all women, though. It is arguably more salient for women from historically under-represented groups, such as SC/STs, who are more exposed to environmental disasters and climatic shocks. This explains our findings regarding the annual expansion of forest cover by female legislators in SC/ST reserved constituencies.

Policy implications

Previous work by Jagnani and Mahadevan (2023) show that female legislators facilitate reduction in air pollution by curbing crop fire incidence. However, unlike pollution, which is often location specific, climate change holds global implications. Hence, climate change mitigation strategies such as forest conservation are likely to generate large-scale positive benefits across national boundaries despite potentially delayed returns. This also implies that climate change abatement measures face the risk of significant underinvestment. Our study demonstrates that the benefit of female political representation is not limited to pollution mitigation, with women politicians better able to balance economic growth and environmental conservation, promoting sustainable development.

Our results are also highly topical in the light of recently passed legislation in India that would reserve one-third of parliamentary and state assembly seats for women in future elections along with the provision of reservation of seats for women from historically marginalised communities. Our findings present the potential for sustained environmental protection on account of mandated political representation of women in higher levels of government in India.

Notes:

The Chipko movement (named after the Hindi word meaning "to stick to") was a forest conservation movement that started in India in the 1970s, where villagers engaged in tree-hugging to prevent deforestation.
The 26th United Nations Climate Change Conference of the Parties (COP-26) in Glasgow is popularly known as “Forest COP”. In the spirit of the COP-26 discussions, the emphasis on forest conservation continued during subsequent COPs.
The delimitation of Indian electoral constituencies was implemented in 2008, with new constituency borders being drawn by the Delimitation Commission. This had an impact on the electoral importance of Indian administrative district, as some districts lost one or several constituencies and others gained some.
In practice, the narrow margin is determined using data-driven algorithms (for example, see Cattaneo et al. 2019)
Densely forested constituencies are considered to be those that belong to the top 25% of the forest cover distribution.