Dear all,

 

I’ve got a couple questions regarding tsls and the automatic specification tests gretl does.

 

First, in a model with k (potentially) endogenous regressors, gretl seems to use k-1 degrees of freedom for the Hausman test rather than k. My impression is that gretl uses the “variable addition” form (eg, Greene p. 323), is that right? If so, shouldn’t the the df be k?

 

The specific model I’m using is from Verbeek’s “schooling” data set. Log wage is the dependent variable, with “black”, “south76” and “smsa76” as exogenous regressors; education, experience and its square (“ed76”, “exp76”, “exp762”) are instrumented regressors. The instruments are age, its square, and a “near college” dummy (plus the 3 exogenous x’s). If I “align” the regressor and instrument lists like this:

 

Model 21: TSLS, using observations 1-3010

Dependent variable: lwage76

Instrumented: ed76 exp76 exp762

Instruments: const black smsa76 south76 nearc4a age76 sq_age76

 

then I don’t get the weak instruments test. But if I switch the order to this:

 

Instruments: const black smsa76 south76 age76 sq_age76 nearc4a

 

I get the test, but with all 0’s for the critical values:

Weak instrument test -

  Cragg-Donald minimum eigenvalue = 1.53611

  Critical values for TSLS bias relative to OLS:

 

     bias       5%      10%      20%      30%

    value     0.00     0.00     0.00     0.00

 

Similar things happen with over-identified models (although in that case I get non-zero critical values). I don’t know as much about these tests as I should, but does the ordering of the instruments really matter?

 

Thanks in advance for any help!

 

PS

 

===============================

Dr. Peter Summers

Assistant Professor

Department of Economics

Texas Tech University

===============================