Using ivreg for timeseries data

[Amina Abubakar]

Hello,

I wanted to confirm whether it is appropriate to use the basic 'ivreg' / 'ivregress' command once you've set your data as a timeseries. Or must it stricitly be 'xtivreg'.

Many thanks,
Amina

[Andrew Musau]

xtset only applies to panel-data estimators and has no bearing on other estimators. However, as with other linear estimators, you can estimate an IV fixed-effects panel data model with ivregress by including indicators, but this is usually not efficient with a large \(N\) dimension.

webuse nlswork, clear
keep in 1/50
xtset id year
xtivreg ln_w age c.age#c.age  (tenure = union south), fe
ivregress 2sls ln_w age c.age#c.age (tenure = union south) i.id, small

Res.:

. xtset id year
       panel variable:  idcode (unbalanced)
        time variable:  year, 68 to 88, but with gaps
                delta:  1 unit

.
. xtivreg ln_w age c.age#c.age  (tenure = union south), fe
note: south omitted because of collinearity

Fixed-effects (within) IV regression            Number of obs     =         35
Group variable: idcode                          Number of groups  =          4

R-sq:                                           Obs per group:
     within  =      .                                         min =          7
     between = 0.6044                                         avg =        8.8
     overall = 0.0135                                         max =         10

                                                Wald chi2(3)      =       7.65
corr(u_i, Xb)  = -0.4440                        Prob > chi2       =     0.1052

------------------------------------------------------------------------------
     ln_wage |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
      tenure |    1.73346   5.927447     0.29   0.770    -9.884122    13.35104
         age |  -1.835107   6.776127    -0.27   0.787    -15.11607    11.44586
             |
 c.age#c.age |   .0253953   .0947635     0.27   0.789    -.1603377    .2111282
             |
       _cons |   27.56593   95.89455     0.29   0.774    -160.3839    215.5158
-------------+----------------------------------------------------------------
     sigma_u |  2.3126893
     sigma_e |   3.997827
         rho |  .25073807   (fraction of variance due to u_i)
------------------------------------------------------------------------------
F  test that all u_i=0:     F(3,28) =     0.05            Prob > F    = 0.9844
------------------------------------------------------------------------------
Instrumented:   tenure
Instruments:    age c.age#c.age union
------------------------------------------------------------------------------

. ivregress 2sls ln_w age c.age#c.age (tenure = union south) i.id, small
note: south omitted because of collinearity


Instrumental variables (2SLS) regression

      Source |       SS       df       MS         Number of obs   =         35
-------------+------------------------------      F(  6,    28)   =       0.04
       Model | -442.730417     6 -73.7884028      Prob > F        =     0.9997
    Residual |  447.513381    28  15.9826207      R-squared       =          .
-------------+------------------------------      Adj R-squared   =          .
       Total |  4.78296402    34  .140675412      Root MSE        =     3.9978

------------------------------------------------------------------------------
     ln_wage |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
      tenure |    1.73346   5.927447     0.29   0.772    -10.40836    13.87528
         age |  -1.835107   6.776127    -0.27   0.789    -15.71537    12.04516
             |
 c.age#c.age |   .0253953   .0947635     0.27   0.791    -.1687189    .2195094
             |
      idcode |
          2  |  -4.211043    12.7215    -0.33   0.743    -30.26985    21.84776
          3  |  -3.676052   10.36385    -0.35   0.725    -24.90543    17.55333
          4  |   .0875298    2.54935     0.03   0.973    -5.134578    5.309637
             |
       _cons |   29.67656   101.5921     0.29   0.772    -178.4255    237.7786
------------------------------------------------------------------------------
Instrumented:  tenure
Instruments:   age c.age#c.age 2.idcode 3.idcode 4.idcode union

.