Gentrification in New York City - Datathon Submission

Last week, I participated in a Datathon organised by Correlation One and Citadel & Citadel Securities, called the Data Open Championship. We spent our evenings during the whole week looking into how to define gentrification, the rough idea that a neighbourhood receives a large influx of high-income residents and departure of the original residents, creating communal tension, and how it impacts noise complaint calls. This is a follow-up work from the popular news article They Played Dominoes Outside Their Apartment For Decades. Then The White People Moved In And Police Started Showing Up.

We used a controlled experiment setup where we were able to estimate the conditional average treatment effect (CATE) on the number of noies complaints. In particular, there were some interesting things to think about about the underlying assumptions, particularly with the potential violation of the Stable Unit Value Treatment Assumption (SUTVA) due to geographical border. Here is the manuscript.