How to test for causality in panel data

[Ahmad Alkhataybeh]

Dears

I am running a logistic regression looking for the probability of an event occurring or not based on a number of explanatory variables.
the form of my model is as follow,

Y= x1 x2 x3 x4
however, the dependent variable is a binary variable (zero,one) and my data set is unbalanced panel.
my question is, how can I make a causality test to investigate if the there as reverse relation between the dependent and the independent variables. Is there is any test similar to the Granger Causality test that can be applied on panel data?

[Carlo Lazzaro]

Ahmad (as per FAQ, please note the strong preference on this forum for family names, too. Thanks)
this Stata thread http://www.stata.com/statalist/archi.../msg00158.html might be helpful.

[Phil Bromiley]

You can estimate this with a two equation model that allows for such reverse causality. I think CMP or the SEM package both allow you to do this.

[Chiraz KARAMTI]

I know my response come very late but it may help other persons who ask about the same thing.
For panel data as M. Baum said it's difficult to have a single result when we have several countries for example. However, recently we have the command "pvargranger" which is applicable for panel data. A very intresting source to understand the theorical background is Hurlin & Mignon (2005): "Une synthèse des tests de racine unitaire sur données de panel", It's in frensh!!