Page 1 of 1

Panel Fixed Effects

Posted: Mon Apr 15, 2019 8:49 am
by carlotacamposc

My name is Carlota. Im currently estimating a panel data with 81 countries from 1998-2017. My subject is about the impact of corruption on poverty. Being poverty my dependent variable, my explanatory variables are corruption (CPI) and education. And my control variables are GDPpc, trade, urban population (...). My panel is unbalanced.

Q1. Since I have T<N (short panel) my question is either I should use cross section fixed or period fixed. Assuming that I should use fixed model rather than a random one.

Q2. After choosing period or cross section fixed in Eviews, which one of the options should I pick: white cross section, white period or white diagonal?

Thanks :)