Dear David,
Thanks for your insights. At other times I have always wondered how stata makes the choice as to which collinear variable to drop, and the "toggling" problem I encountered with -xthtaylor- forced me to confront this question more directly.
Just to clarify, in response to some of your comments below: The different results I get upon running the same -xthtaylor- command come about even though I run no other command inbetween the -xthtaylor- runs.
I have sent you privately my dataset and the specific command. I look forward to your further thoughts upon trying this out on the data.
Best
Hewan
