I have data on intra household consumption of processed food for three time periods. So my time period is 3 and panels are the different household members. My objective is to see the impact of a price increase in the price of processed food on the consumption of different household members. The consumption of processed food is my dependent variable and is expressed in proportions where the sum of all the members equals to 1. As I understand this problem can be modelled through the use of Maarten Buis 's FMLOGIT.
Is it also possible to use FMLOGIT for panel data?
Is it also possible to use FMLOGIT for panel data?
