[Limdep Nlogit List] multivariate tobit

Diana Fletschner fletschn at u.washington.edu
Fri Jun 30 09:38:33 EST 2006


Dear Limdep-users:

I am trying to analyze how spouses who save structure their savings portfolio. That is, what percentage of their savings is in joint accounts, or under his name, or under her name. A basic approach would be to use three Tobit models given that the three dependent variables are bounded between 0 and 100. However, I want to estimate them jointly as a Multivariate Tobit given that the errors are likely to be correlated due to unobserved household characteristics. And, I would like to impose restrictions on the coefficients taking advantage of the fact that we know the sum of the three dependent variables has to equal 100.

Does anyone have and is willing to share a routine to do the Multivariate Tobit (ideally with upper and lower limits and with restrictions) ?  Or any suggestions on how to proceed?  

I would appreciate any help!
Diana


_________________________________
Diana Fletschner
Assistant Professor
Daniel J. Evans School of Public Affairs
University of Washington
(206) 616-1297





More information about the Limdep mailing list