Outreg works with mlogit if you use a version of the dependent variable
that does not include labels for its categories. E.g., I've been
analyzing a DV called "prob" ("principal declared neighborhood problem")
which has categories such as 1=violence/crime, etc. Outreg works with
mlogit in this instance if I use a version of "prob" that does not label
its categories but simply uses 1, 2, 3, 4.
In the meantime, I'll explore estout as well.
