Tobit regression and serial correlation

[Victoria Romanova]

Dear Colleagues,

I ran Tobit regression in STATA and received a comment that there could be a problem with serial correlation. However, I did not use time series; I used data from different countries. How could I solve the problem of serial correlation? Thanks in advance!

[Carlo Lazzaro]

Victoria:
if the reviewer refers to autocorrelation of residuals, you can easily deal it via the -vce(cliuster clusterid)- option for the standard errors available from -tobit-.