Dear friends,

I am running a Fixed Effects regression to test the effect of a policy on firms' patent application. The outcome variable is number of patents.

I first run the simple model with FEs but without any controls:

Code:
Conditional fixed-effects Poisson regression    Number of obs     =     47,536
Group variable: firm_id                         Number of groups  =      7,670

                                                Obs per group:
                                                              min =          2
                                                              avg =        6.2
                                                              max =          9

                                                Wald chi2(9)      =     469.47
Log pseudolikelihood  = -36500.224              Prob > chi2       =     0.0000

                                    (Std. Err. adjusted for clustering on firm_id)
----------------------------------------------------------------------------------
                 |               Robust
 application_num |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------+----------------------------------------------------------------
         treated |  -.0541731   .2172701    -0.25   0.803    -.4800147    .3716685
                 |
application_year |
           1999  |   .5463735   .1321019     4.14   0.000     .2874586    .8052884
           2000  |   1.042662   .2174868     4.79   0.000     .6163956    1.468928
           2001  |    1.35629   .3345565     4.05   0.000     .7005712    2.012008
           2002  |   2.180516   .3422223     6.37   0.000     1.509772    2.851259
           2003  |   2.626014   .3407459     7.71   0.000     1.958164    3.293864
           2004  |   2.883888   .3663026     7.87   0.000     2.165948    3.601827
           2005  |   3.217897   .3785679     8.50   0.000     2.475918    3.959877
           2006  |   3.594803   .4207512     8.54   0.000     2.770146     4.41946
----------------------------------------------------------------------------------
Now the Treated variable is statistically insignificant.
I then add controls:

Code:
Conditional fixed-effects Poisson regression    Number of obs     =     45,948
Group variable: firm_id                         Number of groups  =      7,583

                                                Obs per group:
                                                              min =          2
                                                              avg =        6.1
                                                              max =          9

                                                Wald chi2(13)     =    1555.79
Log pseudolikelihood  = -22604.768              Prob > chi2       =     0.0000

                                    (Std. Err. adjusted for clustering on firm_id)
----------------------------------------------------------------------------------
                 |               Robust
 application_num |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------+----------------------------------------------------------------
         treated |   .1731523   .0770954     2.25   0.025     .0220481    .3242564
total_assets_log |    .324462   .0806509     4.02   0.000     .1663891    .4825349
total_profit_log |  -.6414265   .1401608    -4.58   0.000    -.9161365   -.3667165
  cum_claims_log |   1.192533   .0713113    16.72   0.000     1.052765      1.3323
         age_log |  -.1982098   .1659591    -1.19   0.232    -.5234837    .1270642
                 |
application_year |
           1999  |   .2394592   .0855993     2.80   0.005     .0716876    .4072308
           2000  |   .2442759   .0859471     2.84   0.004     .0758226    .4127292
           2001  |   .1151367   .1044195     1.10   0.270    -.0895217    .3197951
           2002  |   .1297499   .1482964     0.87   0.382    -.1609057    .4204054
           2003  |   .0002627   .1531101     0.00   0.999    -.2998277     .300353
           2004  |  -.2282316   .1756614    -1.30   0.194    -.5725216    .1160585
           2005  |  -.4450677   .2097618    -2.12   0.034    -.8561931   -.0339422
           2006  |  -.7040352   .2517939    -2.80   0.005    -1.197542   -.2105283
----------------------------------------------------------------------------------
As you can see, after adding controls, the Treated variable turns statistically significant.

Can anyone help me understand why this is the case? I know that often time adding controls turns a significant variable insignificant as the controls can absorb some explanation power. But I just can't figure out why in my case, a insignificant variable becomes significant once controls are added.

And which result shall I trust? Does the policy really have a significant impact on firms' patent application?

Thank you very much!