xtbcfe erroe

[Dario Maimone Ansaldo Patti]

Dear All,

I am trying to use the user written routine xtbcfe for bias corrected fixed effect estimation of a panel data model. I installed the routine and attempted to replicate the estimations suggested in the help file.

I got the following error message, which I cannot understand:

. xtbcfe n w wL1 k kL1 kL2 ys ysL1 ysL2, bciters(250) res(iid) ini(bi) lags(2)
command distinct is unrecognized
r(199);

Hence I installed distinct. Then I receive the following error:

 xtbcfe n w wL1 k kL1 kL2 ys ysL1 ysL2, bciters(250) res(iid) ini(bi) lags(2)
               bcfe_ub():  3499  mm_repeat() not found
                 <istmt>:     -  function returned error
r(3499);

As far as you know (if you know the routine), is there any bug in it?

Best

Dario

EDIT

I found the solution in another thread in this forum. But still I have problems with the routine. The results of the estimation (the same suggested in the help file) is reported below:

. xtbcfe n w wL1 k kL1 kL2 ys ysL1 ysL2, bciters(250) res(iid) ini(bi) lags(2) te crit(0.01)


Bootstrap corrected dynamic FE regression       Number of obs      =       751
Group variable : id                             Number of groups   =       140

Resample       : i.i.d.                         Obs per group: min =         5
Initialization : Burn-in                                       avg =       5.4
Convergence    : Yes                                           max =         7


Dependent variable : n
------------------------------------------------------------------------------
                Results                                                      
                 Coefs.  Std. Err.          t      P>|t| [95% Conf.  Interval]
------------------------------------------------------------------------------
L.n           1.0507655          .          .          .          .          .
L2.n         -0.1656689          .          .          .          .          .
w            -0.5516643          .          .          .          .          .
wL1           0.5132341          .          .          .          .          .
k             0.3880882          .          .          .          .          .
kL1          -0.2208651          .          .          .          .          .
kL2          -0.0465805          .          .          .          .          .
ys            0.4757484          .          .          .          .          .
ysL1         -0.7717861          .          .          .          .          .
ysL2          0.1648291          .          .          .          .          .
year4         0.0198841          .          .          .          .          .
year5         0.0347159          .          .          .          .          .
year6         0.0006001          .          .          .          .          .
year7         0.0054584          .          .          .          .          .
year8         0.0242108          .          .          .          .          .
year9         0.0480169          .          .          .          .          .
------------------------------------------------------------------------------
Notes:
- No standard errors computed

Any idea why standard errors are not computed?

[Zubai Khan]

I face the same situation. Have you resolved the problem? If yes kindly reply to this for solution.

[Dario Maimone Ansaldo Patti]

Unfortunately, no. I tried to write to one of the authors of the routine, but no answer received.

[Marco Botta]

By default, XTBCFE does not compute standard errors (see the help file of the command: "choose a method for standard errors and confidence intervals; default is no inference").
You need to use the option "inference" and choose the method you want to adopt in order to obtain standard error estimates.

[Dario Maimone Ansaldo Patti]

Thanks