Estimation of Production function using PRODEST

[Ketan Reddy]

I have estimated my production function using Levinsohn-Petrin and I get insignificant capital, then I run LP using the prodest command and I run the prodest command twice. The first time I get the capital significant at 10% and then at 1%. the coefficients across the three results are similar. so can someone explain why the significance level is changing?

[Sven-Kristjan Bormann]

Have you tried to increase the number of bootstrap repetitions via the reps(#)option? Try different numbers like 10, 50, 100 and then see what changes. The significance levels are changing because the standard errors change. The standard errors are bootstrapped, therefore they may change with every new call of the command.
I have to admit that I don't remember exactly if the random number generator for the bootstrapping process is reset to its previous state whenever the command is called or if the bootstrapping process uses the current state of the random number generator.
If the latter is the case, then it would explain why you see different results on each new call of prodest.

As usual, it would help if you provide an example dataset via the dataex command. Then we could understand maybe better, where your problem is.

[Carlo Lazzaro]

The -seed- option can tame -bootstrap- a bit .

[Ketan Reddy]

@ Seven - Well I did try reps option to increase the number of bootstrapping repetitions. I tried with 10,50,100 and even higher ones and still the results change from time to time.
pasting the result from dataex :

input float(logrealoutput logreallabour logrealcapital logrealfuel)
12.544414 .6931472 16.829235 8.242375
13.277742 .6931472 16.78891 8.808244
13.19438 .6931472 16.760408 8.698816
16.948017 3.433987 14.005565 12.570837
16.898254 6.766191 13.907233 11.88583
16.79255 3.0445225 13.784892 12.517053
16.784952 6.594413 13.561357 12.699282
15.989732 2.397895 13.59584 13.783484
15.856295 2.484907 13.659145 13.479353
16.282335 2.397895 13.565342 13.720532
16.031918 2.484907 13.632194 13.6981
15.33858 5.187386 10.029525 10.15092
15.34944 5.056246 10.02372 10.870413
14.965387 4.919981 9.941513 10.139616
14.630787 4.26268 10.042864 9.893708
17.541351 3.73767 15.881797 14.604465
17.703403 3.8501475 15.868323 14.618804
17.406265 3.8501475 16.566832 14.618473
17.21754 4.0253515 16.74024 14.257714
18.408852 2.70805 15.052722 14.786786
18.539652 2.70805 14.933832 14.618804
18.58492 2.890372 14.927224 14.580732
18.644657 2.484907 15.151998 14.64718
16.894724 2.3025851 14.087758 13.95054
17.084364 2.484907 14.074317 13.328167
17.084682 2.1972246 13.970785 14.15585
17.13053 2.3025851 13.940182 13.778142
18.962921 5.560682 17.111614 13.409267
18.855999 8.078999 16.849113 13.380854
14.209146 1.0986123 9.203898 10.975258
14.494098 1.0986123 9.159636 9.457423
14.672897 1.0986123 10.511754 11.071623
14.589534 1.0986123 10.975606 10.975713
16.51173 1.0986123 13.110142 12.303594
15.982175 1.0986123 13.23336 12.26125
16.425436 .6931472 13.130373 12.122587
16.467236 .6931472 13.040918 12.26996
18.7806 5.247024 15.685203 14.882096
18.507692 5.153292 15.813256 14.21334
18.357534 5.187386 15.687233 14.267076
18.507719 5.204007 15.67212 14.509028
17.296324 2.70805 15.53993 11.950715
17.4814 2.70805 15.425688 13.807874
17.58888 2.995732 15.425488 12.10351
17.416224 2.944439 15.431412 12.05657
17.237667 1.7917595 16.285233 14.330028
17.389746 1.94591 15.985268 14.031017
17.478586 2.1972246 15.999968 14.095225
17.728367 2.0794415 15.92091 13.85225
16.57627 1.609438 13.601822 12.261482
16.585373 1.94591 13.626662 12.06028
16.5506 1.3862944 13.77474 12.664498
17.322903 2.397895 13.427694 12.315246
19.20804 3.912023 19.28263 15.64827
19.914606 4.941642 19.26137 14.41401
19.004162 3.89182 18.849493 15.776983
19.26986 4.65396 19.308134 14.432068
17.02806 3.465736 17.241188 15.512723
17.383762 3.89182 17.213923 15.214787
17.441244 4.248495 16.318312 14.618473
17.66042 4.356709 15.764252 15.58264
14.209146 .6931472 12.114925 12.102003
14.222164 .6931472 12.143933 12.249628
14.410533 1.3862944 11.999567 11.821392
14.42248 .6931472 13.100916 12.302768
17.610344 2.564949 16.349401 13.870496
17.64398 2.564949 16.241121 13.605253
17.907042 2.944439 16.197205 13.558824
17.924482 2.944439 16.127386 13.43952
21.299223 4.317488 19.851046 18.199032
21.724627 6.05444 19.74253 18.1899
21.618393 5.752573 19.68216 18.412712
21.87678 6.089045 19.5763 18.526411
15.392916 1.0986123 13.844273 12.095436
15.278216 2.944439 13.957793 11.849996
15.326824 1.3862944 13.91249 10.900145
20.711435 4.465908 18.349854 16.521387
20.99639 4.812184 18.339144 16.39814
20.80746 4.6151204 18.293386 16.620953
20.87825 4.6728287 18.244783 16.460215
13.720424 1.0986123 12.047035 9.894206
13.84747 1.0986123 12.074222 9.924891
17.061777 3.0445225 15.802552 14.27596
17.44104 2.890372 15.817746 14.095556
17.213894 3.178054 16.077055 14.15585
17.189371 2.772589 16.74375 14.257714
18.745323 4.744932 18.43666 14.752885
19.45594 5.257495 18.29638 14.849327
19.352177 4.6634393 18.320393 14.689932
19.11466 5.375278 18.30153 14.7097
18.39171 4.875197 18.470472 16.17308
19.75814 5.030438 18.418388 16.110458
18.567575 4.89784 18.469646 16.333271
18.66641 5.099866 18.354033 16.080727
18.246332 5.099866 16.520853 13.47341
17.812603 5.129899 16.426826 13.3668
15.989732 3.9512436 13.825536 13.390954
16.237066 4.0943446 13.855594 14.45846
16.202293 3.6635616 14.920538 14.267076
16.118929 3.465736 14.80096 14.046406

[Carlo Lazzaro]

Ketan:
you may consider to report your problem to the Authors of the user-written programme directly, as (quoting from the help file):

Bug reporting

prodest is part of an ongoing project and thus may contain errors and malfunctions. Please submit bugs, comments, and suggestions via email to

[email protected]

Please include steps to reproduce the issue and, if possible, the output of the Stata command creturn list.

[Sven-Kristjan Bormann]

@Ketan: Your example data lacks the necessary IDs for each group and the time variable. Without these two variables, it will not be possible to reproduce your results. Please edit your example data, then it will be possible for us to run the estimations ourselves and it will also help you when writing an email to the authors of this package.