I'm trying to estimate zero inflated count models. The estimation works
rather well until I introduce industry dummies and dummies for the legal
structure of the firms. I then get the following message: "Hessian has
become unstable or asymetric (NC)" (and almost all the dummies are droped
due to colinearity)
Does any of you know what the problem is?
_________________________________________________________________
Vid�oconf�rence plein �cran avec MSN Messenger http://g.msn.fr/FR1001/866
