Investigate causal effects of hydrometeorological extremes on the frequency of protests in India

Extreme hydrometeorological events, resulting from precipitation and temperature anomalies, are projected to increase under climate change, according to studies from the IPCC AR6. The increasingly unstable hydrological environment could be a significant feature acting as a trigger or accelerator for a rising frequency of social unrest events. At this time, however, research on the climate impacts on social unrest, including protest, has not reached any unified conclusions.

In this study, we investigate if there are causal effects between the hydrometeorological anomalies, especially droughts, and the monthly frequency of social protests at the sub-state level in India from 2011 to 2019. The frequency of protests is acquired from the GDELT Project database (https://www.gdeltproject.org/). We use the Standardized Precipitation-Evapotranspiration Index (SPEI) to represent the hydrometeorological anomalies. The confounding variables include climatic, remotely-sensed, and demographic data.

We develop two preliminary models based on Rubin's definition and hypotheses of a causal effect. The pipeline of the first model consists of propensity score matching, hypothesis testing, and regression models to estimate the causal effect. The second model uses the weighted marginal structural model based on the Inverse Probability of Treatment Weight (IPTW). The key idea of the models is to balance the effects of the confounding variables, such as socio-economic development, and hence to quantitatively assess the causal effect of the treatment (extreme SPEI values) on the outcomes (frequency of protests). The results from the model and further analyses are expected to help the researchers in the relevant fields, such as sociologists, select variables for social unrest modeling and provide additional information to monitor and tackle the societal impacts of climate change.