Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • ivreg2/xtivreg2 endog option

    Hello,

    I would like some clarification on the interpretation of tests after running xtivreg2 with the -endog option

    Below is my model & output:

    Code:
    xtivreg2 dly dlpop dlk lly ic ec corr mip gs (dlm lm gsinter corrinter mipinter = onset nofc lonset lnofc ivgsinter1 ivgsinter2 ivcorrinter1
    >   ivcorrinter2 ivmipinter1 ivmipinter2), fe cluster(country_id) small endog(dlm lm gsinter corrinter mipinter)
    
    FIXED EFFECTS ESTIMATION
    ------------------------
    Number of groups =        86                    Obs per group: min =         4
                                                                   avg =      18.6
                                                                   max =        20
    
    IV (2SLS) estimation
    --------------------
    
    Estimates efficient for homoskedasticity only
    Statistics robust to heteroskedasticity and clustering on country_id
    
    Number of clusters (country_id) =     86              Number of obs =     1602
                                                          F( 13,    85) =     4.12
                                                          Prob > F      =   0.0000
    Total (centered) SS     =  1.597127942                Centered R2   =  -0.7060
    Total (uncentered) SS   =  1.597127942                Uncentered R2 =  -0.7060
    Residual SS             =  2.724638077                Root MSE      =   .04258
    
    ------------------------------------------------------------------------------
                 |               Robust
             dly |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
    -------------+----------------------------------------------------------------
             dlm |  -.3289986    .383945    -0.86   0.394    -1.092384    .4343869
              lm |  -.0734162   .0644577    -1.14   0.258    -.2015755     .054743
         gsinter |   .0925139   .1210401     0.76   0.447    -.1481462     .333174
       corrinter |   .1488751   .2384394     0.62   0.534    -.3252063    .6229564
        mipinter |  -.0937414   .1413317    -0.66   0.509    -.3747468    .1872639
           dlpop |  -1.424706   .4682995    -3.04   0.003    -2.355811    -.493601
             dlk |    .026437   .0260477     1.01   0.313    -.0253528    .0782268
             lly |  -.0531175    .028976    -1.83   0.070    -.1107296    .0044945
              ic |  -.0041952   .0018853    -2.23   0.029    -.0079438   -.0004467
              ec |  -.0015734   .0019708    -0.80   0.427    -.0054919    .0023452
            corr |   .0000575   .0083722     0.01   0.995    -.0165887    .0167038
             mip |  -.0008769   .0053712    -0.16   0.871    -.0115562    .0098025
              gs |   -.002366   .0067116    -0.35   0.725    -.0157104    .0109784
    ------------------------------------------------------------------------------
    Underidentification test (Kleibergen-Paap rk LM statistic):              1.530
                                                       Chi-sq(6) P-val =    0.9575
    ------------------------------------------------------------------------------
    Weak identification test (Cragg-Donald Wald F statistic):                0.123
                             (Kleibergen-Paap rk Wald F statistic):          0.159
    Stock-Yogo weak ID test critical values:                       <not available>
    ------------------------------------------------------------------------------
    Hansen J statistic (overidentification test of all instruments):         1.654
                                                       Chi-sq(5) P-val =    0.8946
    -endog- option:
    Endogeneity test of endogenous regressors:                               8.757
                                                       Chi-sq(5) P-val =    0.1192
    Regressors tested:    dlm lm gsinter corrinter mipinter
    ------------------------------------------------------------------------------
    Instrumented:         dlm lm gsinter corrinter mipinter
    Included instruments: dlpop dlk lly ic ec corr mip gs
    Excluded instruments: onset nofc lonset lnofc ivgsinter1 ivgsinter2
                          ivcorrinter1 ivcorrinter2 ivmipinter1 ivmipinter2
    Here for the endogeneity test - I get a p-value of 0.1192, therefore I cannot reject the null hypothesis that
    Under the null hypothesis that the specified endogenous regressors can actually be treated as exogenous, the test statistic is distributed as chi-squared with degrees of freedom equal to the number of regressors tested.
    1. Does this mean that my regressors namely,
    Code:
    dlm lm gsinter corrinter mipinter
    are exogenous and thus there is no need to use instrumental variables or rather that my instruments are suitable.

    For the hansen j statistic -i get a p-value of 0.8946, therefore i cannot reject the null hypothesis that
    The Sargan-Hansen test is a test of overidentifying restrictions. The joint null hypothesis is that the instruments are valid instruments, i.e., uncorrelated with the error term, and that the excluded instruments are correctly excluded from the estimated equation.
    2. Does this mean that my instruments are valid, and not over-identified?

    Thank you,
Working...
X