不是特别明白你的问题，两步法中第一步应该用Xtfrontier命令。

Stata不能直接用命令作一步法随即前沿估计。

Time-invariant inefficiency model               Number of obs      =       493
Group variable: code                            Number of groups   =        29

                                                Obs per group: min =        17
                                                               avg =        17
                                                               max =        17

                                                Wald chi2(2)       =  12334.70
Log likelihood  =  344.30988                    Prob > chi2        =    0.0000

------------------------------------------------------------------------------
         gdp |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         cap |   .7929309   .0089779    88.32   0.000     .7753345    .8105273
         lab |    .358654   .0608994     5.89   0.000     .2392933    .4780147
       _cons |    1.15364   .4426052     2.61   0.009     .2861494     2.02113
-------------+----------------------------------------------------------------
         /mu |   1.282116   .1829066     7.01   0.000     .9236256    1.640607
   /lnsigma2 |  -1.575981   .2654147    -5.94   0.000    -2.096185   -1.055778
  /ilgtgamma |   2.946466   .2874346    10.25   0.000     2.383104    3.509827
-------------+----------------------------------------------------------------
      sigma2 |   .2068045   .0548889                      .1229245    .3479215
       gamma |   .9500962   .0136283                      .9155298    .9709661
    sigma_u2 |   .1964841   .0548896                      .0889025    .3040658
    sigma_v2 |   .0103203    .000678                      .0089915    .0116491
------------------------------------------------------------------------------