Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Ricardo Politi <rbpoliti@hotmail.com> |
To | <statalist@hsphsun2.harvard.edu> |
Subject | st: Number of IV´s - Arellano Bond - GMM |
Date | Tue, 22 Mar 2011 19:44:18 +0000 |
Hi all, I am doing standard tax mimicking econometrics with multiple endogenous regressors. 2SLS static models are not performing well, as Partial R2 and Shea statistics are quite different (also weak IV problems). Dynamic Arellano Bond models are performing better, but I have 312 observations and 47 Instruments (as also only 26 groups). Sargan tests have 22 degrees of freedom. IS there any problem with this last specification? Cuting down the number of IV produce "no sense" results. SO I would like to know if this specification is accurate. Thank you so much, Ricardo Dynamic panel-data estimation, one-step difference GMM Group variable: nestado Number of obs = 312 Time variable : ano Number of groups = 26 Number of instruments = 47 Obs per group: min = 12 F(25, 26) = 653.24 avg = 12.00 Prob > F = 0.000 max = 12 Arellano-Bond test for AR(1) in first differences: z = -2.04 Pr > z = 0.042 Arellano-Bond test for AR(2) in first differences: z = -0.56 Pr > z = 0.577 Sargan test of overid. restrictions: chi2(22) = 30.32 Prob > chi2 = 0.111 (Not robust, but not weakened by many instruments.) Hansen test of overid. restrictions: chi2(22) = 0.00 Prob > chi2 = 1.000 (Robust, but can be weakened by many instruments.) * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/