Dear Statalisters,
I am trying to estimate a fixed effects (conditional) logit model
predicting the probability that a newly hired employee in a given
establishment is white as a function of the hiring manager's race,
conditional on the number of white hires in the establishment. I have
many observations (typically over 100) per store and many positive outcomes
per establishment (also many negative ones). The race of the hiring
manager changes in many, but not all, establishments. There are about 1500
establishments.
After the first iteration, I get log likelihood = -8.988e+307, and stata
stops and prints out an empty table (see output below).
I have had some success running the same regression using smaller
subsamples -- with fewer observations per store -- from the same data set.
I'd greatly appreciate any explanations and/or suggestions.
Thanks very much!
- Laura Giuliano
. clogit white mblack mhisp masian mother t2-t30, group(store)
note: multiple positive outcomes within groups encountered.
note: 114 groups (6827 obs) dropped due to all positive or
all negative outcomes.
Iteration 0: log likelihood = -8.988e+307
Conditional (fixed-effects) logistic regression Number of obs = 253398
LR chi2(33) = .
Log likelihood = . Prob > chi2 = .
------------------------------------------------------------------------------
white | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
mblack | . . . . .
mhisp | . . . . .
masian | . . . . .
mother | . . . . .
t2 | . . . . .
t3 | . . . . .
etc.
---------------------------------------------------
Laura Giuliano
Department of Economics
549 Evans Hall, MC 3880
University of California, Berkeley
phone: (510) 526-9490
email: [email protected]
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
