Hello
I am trying to estimate the effect of bill specific variables (e.g., author, topic) on the choice of a revising legislative chamber (approve; amend; vote to reject; or do nothing). This is for a bicameral country where the revising chamber has those four options (the last two are different because a vote to reject can bring the bill back for a new round of voting).
I estimated models using mlogit and mprobit, which worked very nicely. Tested for iia and did not find a problem. But then I read several criticisms of the iia tests.
I am concerned about it but realized that the suggested model, asmprobit, requires alternative-specific variables, which I do not have.
What is the suggested model in this situation? Should I try mprobit with cmp (my computer has been running such a model for the last few hours and still no results)?
Thanks in advance for your comments.
I am trying to estimate the effect of bill specific variables (e.g., author, topic) on the choice of a revising legislative chamber (approve; amend; vote to reject; or do nothing). This is for a bicameral country where the revising chamber has those four options (the last two are different because a vote to reject can bring the bill back for a new round of voting).
I estimated models using mlogit and mprobit, which worked very nicely. Tested for iia and did not find a problem. But then I read several criticisms of the iia tests.
I am concerned about it but realized that the suggested model, asmprobit, requires alternative-specific variables, which I do not have.
What is the suggested model in this situation? Should I try mprobit with cmp (my computer has been running such a model for the last few hours and still no results)?
Thanks in advance for your comments.
