scpc: Spatial Correlation Robust Inference

This Stata package implements the C-SCPC method described in Müller and Watson (2021) and SCPC method Müller and Watson (2022) for the construction of confidence intervals that account for many forms of spatial correlation. It is implemented as a postestimation command that can be used after the STATA commands "regress", "ivregress", "areg" , "logit" or "probit" as long as these are used with the standard error option "robust" or "cluster". If "cluster" is chosen, then the method assumes that all observations in a cluster are at the same spatial location, and corrects for potential spatial correlation between clusters.

Install

scpc is not currently available from SSC. To install directly from this repository, you can copy and run the following lines in Stata:

// Remove program if it existed previously
cap ado uninstall scpc
// Install most up-to-date version
net install scpc, from("https://raw.githubusercontent.com/ukmueller/SCPC/master/src")

References

Müller, Ulrich K and Mark W. Watson (2021). "Spatial Correlation Robust Inference". Working Paper. https://www.princeton.edu/~umueller/SHAR.pdf.

Müller, Ulrich K and Mark W. Watson (2022). "Spatial Correlation Robust Inference in Linear Regression and Panel Models". Working Paper. https://www.princeton.edu/~umueller/SptialRegression.pdf.