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: tests after xtabond2


From   Carmen Jones <[email protected]>
To   [email protected]
Subject   st: tests after xtabond2
Date   Sun, 13 Oct 2013 13:11:47 -0400

Dear Statalist,

I am struggling to understand a few things, especially results for Hansen tests after running xtabond2. (I did read the help file and how to do xtabond2 article).

I am using the following command:

/xtabond2 Y L.Y L(0/1).(teaching size dsh cmi wage outlier X1 X2 X3) year2007 year2008 year2009 year2010 year2011 year2012, gmm//(L.(Y X1 X2 X3), c) iv( L(0/1).(teaching size dsh//cmi wage outlier)) iv(year2007 year2008 year2009 year2010 year2011 year2012) iv(L.(Z1 Z2 Z3)) robust/

And I get the following test results (data for 6 years; X1, X2 and X3 are endogenous variables; Z1, Z2 and Z3 are additional exogenous instruments; 40 instruments total):

/Instruments for first differences equation
  Standard
    D.(L.Z1 L.Z2 L.Z3)
    D.(year2007 year2008 year2009 year2010 year2011 year2012)
D.(teaching L.teaching size L.size dsh L.dsh cmi L.cmi wage L.wage outlier
    L.outlier)
GMM-type (missing=0, separate instruments for each period unless collapsed)
    L(1/5).(L.Y L.X1 L.X2 L.X3) collapsed
Instruments for levels equation
  Standard
///L.Z1 L.Z2 L.Z3/
    year2007 year2008 year2009 year2010 year2011 year2012
    teaching L.teaching size L.size dsh L.dsh cmi L.cmi wage L.wage outlier
    L.outlier
    _cons
GMM-type (missing=0, separate instruments for each period unless collapsed)
    D.(///L.Y L.X1 L.X2 L.X3/) collapsed
------------------------------------------------------------------------------
Arellano-Bond test for AR(1) in first differences: z = -15.16 Pr > z = 0.000 Arellano-Bond test for AR(2) in first differences: z = 1.15 Pr > z = 0.248
------------------------------------------------------------------------------
Sargan test of overid. restrictions: chi2(16) = 91.42 Prob > chi2 = 0.000
  (Not robust, but not weakened by many instruments.)
Hansen test of overid. restrictions: chi2(16) = 55.95 Prob > chi2 = 0.000
  (Robust, but weakened by many instruments.)

Difference-in-Hansen tests of exogeneity of instrument subsets:
  GMM instruments for levels
Hansen test excluding group: chi2(12) = 51.14 Prob > chi2 = 0.000 Difference (null H = exogenous): chi2(4) = 4.81 Prob > chi2 = 0.307 iv(teaching L.teaching size L.size dsh L.dsh cmi L.cmi wage L.wage outlier L.outlier) Hansen test excluding group: chi2(4) = 26.13 Prob > chi2 = 0.000 Difference (null H = exogenous): chi2(12) = 29.82 Prob > chi2 = 0.003
  iv(year2007 year2008 year2009 year2010 year2011 year2012)
Hansen test excluding group: chi2(12) = 22.29 Prob > chi2 = 0.034 Difference (null H = exogenous): chi2(4) = 33.66 Prob > chi2 = 0.000
  iv(////L.Z1 L.Z2 L.Z3//)
Hansen test excluding group: chi2(13) = 51.51 Prob > chi2 = 0.000 Difference (null H = exogenous): chi2(3) = 4.44 Prob > chi2 = 0.218

/My questions are:

1) What does the Hansen test tell me? I have tried adding more lags but the p-value remains close to 0 and I only have 6 periods available. Is there anything I can do about it? Should I care about it? 2) Looking at the Hansen and difference tests for the different groups of variables, it seems that the year dummies, for example, are a problem. How do I change this given that time dummies are important to include if I understand correctly?

Overall, I am trying to understand what I am doing wrong.

Thank you!

Carmen




*
*   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