st: Obtaining fixed effects using the predict command ; sorting issues
Fri, 19 Aug 2005 10:28:22 +0100 (BST)
To Whom It May Concern,
I am estimating a fixed-effects model on a dynamic panel using the xtreg, fe command. I would like to obtain an estimate of the fixed effects. When I use the predict command (with option ,u for fixed effects) I get an error message telling me the data are not sorted. However, simply sorting the data by panel-variable and time-variable does not resolve the issue. My question is; what am I doing wrong? Or; how do I sort my data?
PS Obviously, I could compute the fixed effects in a more roundabout fashion by calculating fitted values (the predict command does enable me to obtain fitted values), and computing the mean residual, but I have the gut feeling there must be an easier way of obtaining the fixed effects.
* For searches and help try: