Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: Fixed effects vector decomposition

From   Nick Cox <>
Subject   Re: st: Fixed effects vector decomposition
Date   Mon, 1 Oct 2012 09:28:22 +0100


and its references before proceeding any further.


On Mon, Oct 1, 2012 at 8:39 AM, fernando luiz mistura
<> wrote:
> I am testing the effect of a rarely-changing variable (FDI restrictions) lon the level of FDI stocks.
> Under OLS, Random Effects, the effect is significant. However, with Fixed effects it is not. Hausman test tells me that Fixed Effects would be more appropriate. But isn't Fixed Effects unlikely to be adaptable to variable that change slowly overtime or not change at all for some countries in the panel.
> Wityh Fixed Effects Vector Decomposition, I significant coefficients but they are much higher than with OLS RE or FE. Is this normal?

*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index