Bookmark and Share

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


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

st: Estimating a fractional logit with fixed effects in a non-stationary panel context


From   Jonathan Argent <[email protected]>
To   [email protected]
Subject   st: Estimating a fractional logit with fixed effects in a non-stationary panel context
Date   Mon, 24 Feb 2014 16:59:57 +0200

I would like to estimate the impact of market structure on technology
diffusion in the telecoms industry using a quarterly panel (56
quarters) of all countries (over 200).  I believe that the most
appropriate model for this is a fractional logit with fixed effects
(using the glogit command, or gls, family(binomial) link(logit), with
dummies for different market structure on the right hand side (and
using the margins command to interpret).

Papke and Wooldridge (2008) suggest that in the context of obvserving
the whole population (as I do) the fractional logit with firm specific
intercepts allows unobserved heterogeneity to enter in a flexible way,
but also acknowledges that the fixed effects estimates may be
inconsistent with small T.

1) Are 56 quarters enough that the fixed effects estimation will be consistent?
2) Obviously my data are non-stationary - but I understand that in a
panel context, this is often ok. Would this present a problem for my
standard errors? And if so, how might I correct them?
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/


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