Hi dear profs and colleagues,
I am running this model. In Reg estimation, there is no collinearity but when I run fixed effect, the main explanatory variable is omitted due to collinearity. I tried to drop the control variables and estimate only the dependent variable and predictor, though collinearity still exists and eliminates the independent variable. Could you please share your ideas in this regard? Thanks.
I am running this model. In Reg estimation, there is no collinearity but when I run fixed effect, the main explanatory variable is omitted due to collinearity. I tried to drop the control variables and estimate only the dependent variable and predictor, though collinearity still exists and eliminates the independent variable. Could you please share your ideas in this regard? Thanks.
Code:
clear
input double(year NPC_FIC firm_age) float(region mig_firm immi_sh foreign_aff sector)
2010 500000001 7 6 0 .012691068 . .
2010 500000002 25 2 0 .035209775 . .
2011 500000002 26 2 1 .035209775 . .
2012 500000002 27 2 1 .035209775 . .
2013 500000002 28 2 0 .035209775 . .
2014 500000002 29 2 0 .035209775 . .
2015 500000002 30 2 0 .035209775 . .
2016 500000002 31 2 0 .035209775 . .
2017 500000002 32 2 0 .035209775 . .
2018 500000002 33 2 0 .035209775 . .
2010 500000033 25 1 0 .017485183 . .
2011 500000033 26 1 0 .017485183 . .
2012 500000033 27 1 0 .017485183 . .
2013 500000033 28 1 0 .017485183 . .
2014 500000033 29 1 0 .017485183 . .
2015 500000033 30 1 0 .017485183 . .
2016 500000033 31 1 0 .017485183 . .
2017 500000033 32 1 0 .017485183 . .
2018 500000033 33 1 0 .017485183 . .
2019 500000033 34 1 0 .017485183 . .
2020 500000033 35 1 0 .017485183 . .
2010 500000050 25 2 0 .035209775 . .
2011 500000050 26 2 0 .035209775 . .
2012 500000050 27 2 0 .035209775 . .
2013 500000050 28 2 0 .035209775 . .
2014 500000050 29 2 0 .035209775 . .
2015 500000050 30 2 0 .035209775 . .
2010 500000083 5 2 0 .035209775 . .
2011 500000083 6 2 0 .035209775 . .
2012 500000083 7 2 0 .035209775 . .
2013 500000083 8 2 0 .035209775 . .
2014 500000083 9 2 0 .035209775 . .
reg mig_jump immi_sh firm_age foreign_aff i.sector i.region, vce (robust)
Linear regression Number of obs = 108,989
F(16, 108972) = 20.92
Prob > F = 0.0000
R-squared = 0.0032
Root MSE = .15762
------------------------------------------------------------------------------
| Robust
mig_jump | Coefficient std. err. t P>|t| [95% conf. interval]
-------------+----------------------------------------------------------------
immi_sh | .2279302 .1819923 1.25 0.210 -.1287722 .5846326
firm_age | .0001614 .0000319 5.06 0.000 .0000989 .0002239
foreign_aff | -.0086047 .0026538 -3.24 0.001 -.0138061 -.0034033
|
sector |
6 | -.0067889 .0021552 -3.15 0.002 -.0110131 -.0025647
7 | -.0144425 .0012577 -11.48 0.000 -.0169075 -.0119774
9 | .0015004 .001927 0.78 0.436 -.0022765 .0052774
10 | .0043284 .0043409 1.00 0.319 -.0041796 .0128365
11 | -.0080704 .0033 -2.45 0.014 -.0145384 -.0016024
12 | -.0112652 .001748 -6.44 0.000 -.0146913 -.0078391
13 | -.0107205 .0026038 -4.12 0.000 -.0158239 -.0056171
|
region |
2 | 4.04e-06 .0034589 0.00 0.999 -.0067753 .0067834
3 | -.0052456 .0095821 -0.55 0.584 -.0240264 .0135351
4 | -.0148905 .0258107 -0.58 0.564 -.0654792 .0356982
5 | -.0067527 .0074196 -0.91 0.363 -.021295 .0077896
6 | -.0005219 .0034816 -0.15 0.881 -.0073458 .006302
7 | -.0051084 .0033725 -1.51 0.130 -.0117185 .0015017
|
_cons | .0303745 .0042306 7.18 0.000 .0220825 .0386665
------------------------------------------------------------------------------
.
end of do-file
. do "C:\Users\35193\AppData\Local\Temp\STD52d8_000000.tmp"
. xtreg mig_jump immi_sh firm_age foreign_aff i.sector i.region i.year , fe cluster (NPC_FIC )
note: immi_sh omitted because of collinearity.
Fixed-effects (within) regression Number of obs = 108,989
Group variable: NPC_FIC Number of groups = 23,475
R-squared: Obs per group:
Within = 0.0086 min = 1
Between = 0.0061 avg = 4.6
Overall = 0.0033 max = 10
F(24,23474) = 88.27
corr(u_i, Xb) = -0.2588 Prob > F = 0.0000
(Std. err. adjusted for 23,475 clusters in NPC_FIC)
------------------------------------------------------------------------------
| Robust
mig_jump | Coefficient std. err. t P>|t| [95% conf. interval]
-------------+----------------------------------------------------------------
immi_sh | 0 (omitted)
firm_age | -.0002446 .0008499 -0.29 0.774 -.0019104 .0014213
foreign_aff | -.0007732 .0186608 -0.04 0.967 -.0373497 .0358032
|
sector |
6 | -.0375509 .0225933 -1.66 0.097 -.0818352 .0067333
7 | -.0307865 .0121707 -2.53 0.011 -.0546418 -.0069311
9 | -.0352974 .0236857 -1.49 0.136 -.081723 .0111282
10 | -.0371334 .0131044 -2.83 0.005 -.0628188 -.0114479
11 | -.0253307 .0261081 -0.97 0.332 -.0765042 .0258429
12 | -.0223872 .0218179 -1.03 0.305 -.0651517 .0203773
13 | -.0207113 .0262835 -0.79 0.431 -.0722287 .030806
|
region |
2 | -.0128642 .0165671 -0.78 0.437 -.0453367 .0196083
3 | -.0334628 .0172836 -1.94 0.053 -.0673398 .0004143
4 | -.0069956 .0381937 -0.18 0.855 -.0818577 .0678665
5 | -.0265096 .0252737 -1.05 0.294 -.0760476 .0230285
6 | -.0111455 .047931 -0.23 0.816 -.1050933 .0828023
7 | .0407561 .0680305 0.60 0.549 -.0925881 .1741002
|
year |
2011 | .0365941 .0017047 21.47 0.000 .0332528 .0399353
2012 | .0235153 .0022009 10.68 0.000 .0192015 .0278292
2013 | .0218529 .0029972 7.29 0.000 .0159782 .0277277
2014 | .0272034 .0038541 7.06 0.000 .0196492 .0347576
2015 | .0267968 .0046023 5.82 0.000 .017776 .0358176
2016 | .0279896 .0054554 5.13 0.000 .0172966 .0386826
2017 | .0320082 .0063008 5.08 0.000 .0196583 .0443581
2018 | .0458582 .0073004 6.28 0.000 .031549 .0601674
2019 | .0532852 .0081849 6.51 0.000 .0372424 .0693281
|
_cons | .0408039 .0271331 1.50 0.133 -.0123788 .0939866
-------------+----------------------------------------------------------------
sigma_u | .06500605
sigma_e | .16249555
rho | .1379597 (fraction of variance due to u_i)
------------------------------------------------------------------------------
Comment