[Limdep Nlogit List] unbalanced panel data with GMM estimator

[Anat Manes]

Dear Limdep users,


We are considering using the NLSUR- GMM procedure with an unbalanced panel
data.

We have four groups (firms) of observations; each includes a different
number of observations, (each corresponds to a different time period).
Moreover, there are cases in which time periods are not successive.
For example: group 1 includes observations of periods: 2-4, 7-18, 23-30,
group 2, on the other hand, includes observations of periods: 1-3, 5-9,
12-18, 22-30 (and the same principal applies to groups 3 and 4).

We would like to know how to instruct NLOGIT to attribute the right group
and time period to each observation line. (In the example above, it should
refer to the sixth observation in groups 1 and 2, as periods 9 and 7
respectively)

We would be grateful for any assistance.


Anat Tchetchik,

The Department of Agricultural Economics
The Hebrew University, Rehovot 76100 Israel
Phone: 972-8-9489231
Fax: 972-8-9466267