Dear all,

I start my project with panel data and xtlogit command.

In the case of presence of serial autocorrelation, robust-VCE is a good solution. However, in the case of presence of cross-sectional autocorrelation, how should I address it?