Hi all,

I have some questions regardings the dynamic random-effects probit model, especially with the command -xtdyn-

So my panel data has information on 2001, 2003, 2005, 2007, 2009, 2011 and 2013. When I try -xtdyn-, it returns the error saying "no observations r(2000)". I thought the reason for that error is that the years are not consecutive.

So to test my thought, I re-defined the year variable to make it consecutive. So instead of 2001, 2003, 2005, 2007, 2009, 2011 and 2013, it is now 2001, 2002, 2003, 2004, 2005, 2006 and 2007. And the command works.

I suspect that redefining the year variable is going to create some problems, but I am not so sure. Is there going to be any problem if I do that? If so, could you please let me know what the problem will be?

The dynamic random-effects probit model uses the lagged and initial value of the dependent variable and initial values and within-unit averages of the time-varying explanatory variables. Can I construct those variables manually and use -xtprobit-? Would they be equivalent?

Thank you for your help.