Causal Inference with Spatio-Temporal Data: Estimating the Effects of Airstrikes on Insurgent Violence in Iraq

Abstract

Many causal processes have spatial and temporal dimensions. Yet the classic causal inference framework is not directly applicable when the treatment and outcome variables are generated by spatio‐temporal point processes. We extend the potential outcomes framework to these settings by formulating the treatment point process as a stochastic intervention. Our causal estimands include the expected number of outcome events in a specified area under a particular stochastic treatment assignment strategy. Our methodology allows for arbitrary patterns of spatial spillover and temporal carryover effects. Using martingale theory, we show that the proposed estimator is consistent and asymptotically normal as the number of time periods increases. We propose a sensitivity analysis for the possible existence of unmeasured confounders, and extend it to the Hájek estimator. Simulation studies are conducted to examine the estimators' finite sample performance. Finally, we illustrate the proposed methods by estimating the effects of American airstrikes on insurgent violence in Iraq from February 2007 to July 2008. Our analysis suggests that increasing the average number of daily airstrikes for up to 1 month may result in more insurgent attacks. We also find some evidence that airstrikes can displace attacks from Baghdad to new locations up to 400 km away.

Text-Transport: Toward Learning Causal Effects of Natural Language

“…These works aim to understand the conditions under which causal effects estimated on specific data can generalize or be transported to broader settings (Tipton, 2014;Bareinboim and Pearl, 2021). Prior work has also used density ratio-style importance weights to estimate average causal effects with high-dimensional interventions (de la Cuesta et al, 2022;Papadogeorgou et al, 2022).…”

Planetary Causal Inference: Implications for the Geography of Poverty

“…While spatial regression remains a dominant form of statistical analysis with geographic data, we note the emergence of some more recent methodologies that analyze the effect of spatial treatments using stochastic interventions(Papadogeorgou et al, 2022), somewhat complicating the dichotomy between "inflexible" statistical and "flexible" machine learning models.…”

From Victims to Dissidents: Legacies of Violence and Popular Mobilization in Iraq (2003–2018)

“…In TablesA13 and A14in Appendix 10 of the Supplementary Material, we run models with an alternative measure of underlying conflict events: the number of U.S. airstrikes in a district during the war (sourced fromPapadogeorgou et al 2022). Results are similar when including this alternative conflict control.12 Ideally, we would also control for levels of protest in districts during the period before the U.S. invasion (i.e., the pretreatment period).…”