I am running a regression with 1 million observation
outcome: occupation choice(5 category)
fixed effects: high school fixed effects(about 1000 groups), college fixed effects(about 800 groups), and neighborhood fixed effects (about 600 groups)
no matter I use
xtset school
xtmlogit occupation age gender year2000-year2008 i.college i.neighborhood, fe vce(cluster school)
or
mlogit occupation age gender year2000-year2008 i.school i.college i.neighborhood, fe vce(cluster school)
in stata, the computational process is too slow what i could not get results (no error reported, either)
Is there any alternative method for this issue?
alternative method for fast estimation
New contributor
Qiuyi Wang is a new contributor to this site. Take care in asking for clarification, commenting, and answering.
Check out our Code of Conduct.