We propose a spatiotemporal causal inference framework that fully leverages microlevel, granular data. ATE, heterogeneity, and mediation — all in one framework. Now with updated results and visualizations!

New working paper: “Survey Estimates of Wartime Mortality,” with Gary King, available at gking.harvard.edu/sibs. We provide the first formal proofs of the statistical properties of existing mortality estimators, along with empirical illustrations, to develop intuitions that guide best practices.

Excited to present our poster on spatiotemporal causal inference at #PolMeth 2025. Looking forward to seeing many of you there! Paper: arxiv.org/abs/2504.03464 Package: github.com/mmukaigawara...

Interested in causal inference using high-frequency, fine-grained geospatial data? Check out our 2025 PolMeth poster on spatial-temporal causal inference, designed by the wildly talented @mitsurumu.bsky.social. We examine the effects of US airstrikes and civilian harm on insurgent attacks in Iraq

New paper alert (hey, I can't doom scroll all the time): This one's on doing causal inference with "microlevel data" where we suspect that the treatment has spatial spillover & temporal carryover effects. We illustrate our new approach + package w/ application to US counterinsurgency efforts in Iraq

How can we identify causal effects using micro-level data? Our new framework estimates ATEs, probes causal mechanisms, and uncovers heterogeneity—all in one. We illustrate it with an analysis of airstrikes and insurgent attacks in Iraq. arxiv.org/abs/2504.03464

Here's one example of how we used geocausal to estimate the effects of different distributions of US airstrikes in Iraq on insurgent attacks (open access): academic.oup.com/jrsssb/artic...

Want to do causal inference using high-frequency geospatial data with temporal carryover or spatial spillover effects? We've created a new R package, geocausal, that lets you estimate counterfactuals at user-specified intervals for different distributions of a treatment github.com/mmukaigawara...

New in AJPS with @carlynwayne.bsky.social @mitsurumu.bsky.social @profmholmes.bsky.social: how do group dynamics affect assessments of resolve and costly signals? #polisky onlinelibrary.wiley.com/doi/10.1111/...

Interested in doing causal inference with spatio-temporal data? We've got a new R package, geocausal, that allows you to estimate causal effects + counterfactuals for super fine grained data over user-specified spatial + temporal windows. Download it here. (Paper TK soon). polisky