save coefficients of interaction term

[Minnie Smith]

Hi, I have a regression of the following form:

 ivreg2 y (c.x##i.year = c.w##i.year) i.year

Could someone please help me in creating a variable that stores the coefficient from this term? I'm running this regression in a loop for different y values so need to store each coefficient.

Thanks!

[William Lisowski]

I'm not quite sure precisely what you are looking for, but he output of help ivreg2 tells us that the command stores its coefficients in the Stata matrix e(b), so here is some example code for retrieving coefficients corresponding to those in the ivreg2 report. If you're not familiar with Stata matrix commands, start with the output of help matrix. If you're new to returned estimations, start with the output of help ereturn.

. use http://fmwww.bc.edu/ec-p/data/hayashi/griliches76.dta, clear
(Wages of Very Young Men, Zvi Griliches, J.Pol.Ec. 1976)

. ivreg2 lw s  i.year (c.s##i.year = c.rns##i.year)
Warning - duplicate variables detected
Duplicates:         s 67.year 68.year 69.year 70.year 71.year 73.year

IV (2SLS) estimation
--------------------

Estimates efficient for homoskedasticity only
Statistics consistent for homoskedasticity only

                                                      Number of obs =      758
                                                      F( 13,   744) =     9.39
                                                      Prob > F      =   0.0000
Total (centered) SS     =  139.2861498                Centered R2   =  -1.1211
Total (uncentered) SS   =  24652.24662                Uncentered R2 =   0.9880
Residual SS             =  295.4456675                Root MSE      =    .6243

------------------------------------------------------------------------------
          lw |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
    year#c.s |
         67  |  -.4913112   1.540141    -0.32   0.750    -3.509933     2.52731
         68  |  -.3475382   .2634286    -1.32   0.187    -.8638487    .1687723
         69  |  -.1369025   .3759501    -0.36   0.716    -.8737511    .5999461
         70  |  -.6805452   .5553787    -1.23   0.220    -1.769067    .4079769
         71  |  -.7856451   .4655452    -1.69   0.091    -1.698097    .1268068
         73  |  -.5531592    .532216    -1.04   0.299    -1.596283     .489965
             |
           s |   .4887797   .2391004     2.04   0.041     .0201516    .9574078
             |
        year |
         67  |   5.849261   19.34329     0.30   0.762     -32.0629    43.76142
         68  |   4.112184   3.214649     1.28   0.201    -2.188412    10.41278
         69  |   1.387305   4.828885     0.29   0.774    -8.077135    10.85175
         70  |   8.878234   7.537416     1.18   0.239     -5.89483     23.6513
         71  |   10.40018   6.257345     1.66   0.096    -1.863987    22.66435
         73  |    7.39559    7.74392     0.96   0.340    -7.782214    22.57339
             |
       _cons |  -.4039679   2.881636    -0.14   0.889    -6.051871    5.243936
------------------------------------------------------------------------------
Underidentification test (Anderson canon. corr. LM statistic):           0.561
                                                   Chi-sq(2) P-val =    0.7555
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):                0.079
Stock-Yogo weak ID test critical values:                       <not available>
------------------------------------------------------------------------------
Sargan statistic (overidentification test of all instruments):           4.720
                                                   Chi-sq(1) P-val =    0.0298
------------------------------------------------------------------------------
Instrumented:         67.year#c.s 68.year#c.s 69.year#c.s 70.year#c.s
                      71.year#c.s 73.year#c.s
Included instruments: s 67.year 68.year 69.year 70.year 71.year 73.year
Excluded instruments: rns 67.year#c.rns 68.year#c.rns 69.year#c.rns
                      70.year#c.rns 71.year#c.rns 73.year#c.rns
Duplicates:           s 67.year 68.year 69.year 70.year 71.year 73.year
------------------------------------------------------------------------------

.
. matrix coeff = e(b)

. matrix list coeff

coeff[1,16]
      66b.year#    67.year#    68.year#    69.year#    70.year#    71.year#    73.year#
          co.s         c.s         c.s         c.s         c.s         c.s         c.s
y1           0  -.49131124  -.34753817  -.13690248  -.68054523  -.78564513   -.5531592

                       66b.         67.         68.         69.         70.         71.
             s        year        year        year        year        year        year
y1   .48877974           0   5.8492611    4.112184   1.3873052   8.8782343   10.400184

            73.            
          year       _cons
y1   7.3955899  -.40396787

. display coeff[1,2]
-.49131124

. display coeff[1,colnumb(coeff,"67.year#c.s")]
-.49131124