Difference-in-Differences with Geocoded Microdata

TL;DR

This paper develops a new nonparametric method for estimating how treatment effects vary with distance using geocoded microdata, improving accuracy over traditional fixed-ring approaches.

Contribution

It introduces a flexible estimator that captures the decay of treatment effects over distance, relaxing previous assumptions and enhancing analysis of spatial treatment impacts.

Findings

Traditional methods underestimate effects near treatment zones.

The new estimator accurately captures the treatment effect curve.

Application shows previous studies may misestimate effect sizes and ranges.

Abstract

This paper formalizes a common approach for estimating effects of treatment at a specific location using geocoded microdata. This estimator compares units immediately next to treatment (an inner-ring) to units just slightly further away (an outer-ring). I introduce intuitive assumptions needed to identify the average treatment effect among the affected units and illustrates pitfalls that occur when these assumptions fail. Since one of these assumptions requires knowledge of exactly how far treatment effects are experienced, I propose a new method that relaxes this assumption and allows for nonparametric estimation using partitioning-based least squares developed in Cattaneo et. al. (2019). Since treatment effects typically decay/change over distance, this estimator improves analysis by estimating a treatment effect curve as a function of distance from treatment. This is contrast to the traditional method which, at best, identifies the average effect of treatment. To illustrate the advantages of this method, I show that Linden and Rockoff (2008) under estimate the effects of increased crime risk on home values closest to the treatment and overestimate how far the effects extend by selecting a treatment ring that is too wide.