dynamic multinomial logit

[Jimmy Esc]

I want to estimate a dynamic multinomial logit model to observe job mobility, i have 4 categories: 1. employed with low wage, 2 employed with high wage 3 unemployed and 4. No participing in labor face. The panel database is something like the following

id (individual identifier)	year	age	education (years of school)	status laboral
1	78	29	7	1
1	79	30	7	2
1	80	31	7	1
2	79	45	10	1
....	...	..	..	.,,
...	..	..	..	...

Note: of course i have more explanatary variables

I know that the command to use in this type of estimates is gsem, I read the manual but I don´t quite understand how it works. For example ->
http://www.stata.com/stata-news/news29-2/xtmlogit/
here Rebbeca Pope mention something similar to what I am looking to do and how to use the gsem command, But I do not understand why the database that uses the example does not mention the variable time??

I would also like to know which tools have stata to analyze this type of models

Note: i work in stata 13


thanks and regards

[Stephen Jenkins]

See also perhaps the article by Hahn and Uhlendorff in the Stata Journal 6(2), 2006 -- free download

[Jorge Duarte-Rolon]

Hi Jimmy,

I have exactly the same concern as yours. Have you managed to solve it?

Thanks